Machine learning classification of solid tumors based on gene expression

Abstract

The present disclosure provides systems and methods for machine learning classification of solid tumor based on gene expression data and optionally clinical characteristics data. The method can include: a) obtaining a data set containing gene expression measurements of a biological sample from a patient of at least two lung disease-associated genes selected from the gene set capable of classifying a solid tumor as benign or malignant, and optionally clinical characteristics data of one or more clinical characteristics of the patient; b) providing the data set as an input to a machine-learning model trained to generate an inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor; c) receiving, as an output of the machine-learning model, the inference indicating whether the data set is indicative of the malignant solid tumor or the benign solid tumor; and d) electronically outputting a report classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor.

Description

Attorney Docket No.225234-718601 / PCT MACHINE LEARNING CLASSIFICATION OF SOLID TUMORS BASED ON GENE EXPRESSION CROSS-REFERENCE

[0001] This application claims the benefit of priority to U.S. Provisional Application No.63 / 539,073, filed on September 18, 2023, which is incorporated herein by reference in its entirety. BACKGROUND

[0002] Early diagnosis of cancer can be critical for successful treatment. Often cancer diagnosis requires biopsy of a solid tumor to determine whether it is malignant or benign. Certain tumors including lung, pancreatic, ovarian, kidney and brain, require collecting biopsy samples from the patients through invasive and / or painful surgery. The risks of such procedures include tissue and organ damage. For example, brain tumor biopsy can cause brain injury due to the removal of brain tissue, nerve damage from surgical tools used during tissue removal, and seizures due to the resulting scars left on the brain. Lung nodules are common, often detected in screenings of patients experiencing no symptoms of lung disease. Among subjects having lung nodules, only a fraction are eventually diagnosed with a cancer. Noncancerous causes of lung nodules can include e.g., mycobacterial or fungal infection, autoimmune diseases, air pollutants, and scarring from previous insult. Large lung nodules typically warrant an invasive biopsy or removal by thoracic surgery. The percentage of lung nodules eventually identified as cancerous has been estimated to be as low as 40%. Given the potential harm of biopsy or thoracic surgery, less invasive testing for lung cancer is needed. A simple noninvasive test for such solid tumors, e.g., a blood test, would greatly reduce both the potential for patient harm and medical costs. SUMMARY

[0003] The present disclosure provides systems and methods for assessing a solid tumor of a patient using machine learning and uses thereof. In some embodiments, methods and systems for assessing a solid tumor of a patient by classifying gene expression with machine learning. Certain Embodiments

[0004] The present disclosure provides systems and methods for assessing a solid tumor of a patient using machine learning and uses thereof.

[0005] Provided herein are methods for determining a gene set capable of classifying a solid tumor as benign or malignant without biopsy, the methods comprising: a) obtaining a reference data set comprising a plurality of individual reference data sets, wherein a respective individual reference data set of the plurality of individual reference data sets comprises i) gene expression measurements of a plurality of genes of a reference biological sample from a reference subject having a reference solid tumor, ii)Attorney Docket No.225234-718601 / PCT data regarding whether the reference solid tumor of the reference subject is benign or malignant, and iii) optionally clinical characteristics data of one or more clinical characteristics of the reference subject, and, wherein the reference biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; b) training a machine learning model using the reference data set, wherein the machine learning model is trained to infer whether a solid tumor is benign or malignant based at least in part on gene expression measurements of the plurality of genes, and optionally clinical characteristics data of the one or more clinical characteristics; c) determining feature importance values of the plurality of genes; and d) determining the gene set based at least in part on the feature importance values.

[0006] In some embodiments, the solid tumor is: (i) a pancreatic tumor, an ovarian tumor, a kidney tumor, or a brain tumor; (ii) a pancreatic tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to pancreatic cancer; (iii) an ovarian tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to ovarian cancer; (iv) a brain tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to brain cancer; (v) a kidney tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to kidney cancer; (vi) a pancreatic tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to pancreatic cancer; (vii) an ovarian tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to ovarian cancer; (viii) a brain tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to brain cancer; or (ix) a kidney tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to kidney cancer.

[0007] Provided herein are methods for developing a trained machine learning model capable of classifying a solid tumor of a patient as benign or malignant, the method comprising: (a) obtaining a first reference data set comprising a plurality of first individual reference data sets, wherein a respective first individual reference data set of the plurality of first individual reference data sets comprises i) gene expression measurements of a plurality of genes of a reference biological sample from a reference subject having a reference solid tumor, ii) data regarding whether the reference solid tumor is benign or malignant, and iii) optionally clinical characteristics data of one or more clinical characteristics of the reference subject, and wherein the reference biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; (b) training a first machine learning model using the first reference data set, wherein the first machine learning model is trained to infer whether a solid tumor is benign or malignant,Attorney Docket No.225234-718601 / PCT based at least in part on one or more predictors selected from the plurality of genes, and optionally the one or more clinical characteristics; (c) determining feature importance values of the one or more predictors of the first machine learning model; (d) selecting A predictors of the first machine learning model based at least in part on the feature importance values, wherein A is an integer from 5 to 2000; and (e) training a second machine learning model based at least in part on a second reference data set comprising a plurality of second individual reference data sets, wherein a respective second individual reference data set of the plurality of second individual reference data sets comprises i) measurement data of the A predictors of the reference subject, and ii) data regarding whether the reference solid tumor of the reference subject is benign or malignant, to obtain the trained machine learning model, wherein the trained machine learning model is trained to infer whether a solid tumor is benign or malignant, based at least in part on measurement data of the A predictors.

[0008] In some embodiments, the solid tumor is: (i) a pancreatic tumor, an ovarian tumor, a kidney tumor, or a brain tumor; (ii) a pancreatic tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to pancreatic cancer; (iii) an ovarian tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to ovarian cancer; (iv) a brain tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to brain cancer; (v) a kidney tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to kidney cancer; (vi) a pancreatic tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to pancreatic cancer; (vii) an ovarian tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to ovarian cancer; (viii) a brain tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to brain cancer; (ix) a kidney tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to kidney cancer.

[0009] In some embodiments, the A predictors have top 5 to 200 feature importance values.

[0010] In some embodiments, the trained machine learning model has: (i) an accuracy of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (ii) a sensitivity at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iii) a specificity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iv) a positive predictiveAttorney Docket No.225234-718601 / PCT value at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (v) a negative predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; or (vi) a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99.

[0011] In some embodiments, the first machine learning model and second machine learning model is independently trained using a linear regression, a logistic regression, a Ridge regression, a Lasso regression, an elastic net (EN) regression, a support vector machine (SVM), a gradient boosted machine (GBM), a k nearest neighbors (kNN), a generalized linear model (GLM), a naïve Bayes (NB) classifier, a neural network, a Random Forest (RF), a deep learning algorithm, a linear discriminant analysis (LDA), a decision tree learning (DTREE), an adaptive boosting (ADB), or any combination thereof.

[0012] Provided herein are methods for assessing a solid tumor of a patient, the method comprising: a) obtaining a data set comprising i) gene expression measurements of a biological sample from the patient, of at least 2 genes selected from the gene set described herein, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient, wherein the biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; b) providing the data set as an input to a machine-learning model trained to generate an inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor; c) receiving, as an output of the machine-learning model, the inference indicating whether the data set is indicative of the malignant solid tumor or the benign solid tumor; and d) electronically outputting a report classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor.

[0013] In some embodiments, the solid tumor is: (i) a pancreatic tumor, an ovarian tumor, a kidney tumor, or a brain tumor.

[0014] In some embodiments, the solid tumor is a pancreatic tumor, and the at least 2 genes are selected from the gene set described herein, wherein the gene set is capable of classifying the pancreatic tumor as benign or malignant; and optionally wherein the one or more clinical characteristics are selected from a group of clinical characteristics related to pancreatic cancer.Attorney Docket No.225234-718601 / PCT

[0015] In some embodiments, the solid tumor is an ovarian tumor, and the at least 2 genes are selected from the gene set described herein, wherein the gene set is capable of classifying the ovarian tumor as benign or malignant; and optionally wherein the one or more clinical characteristics are selected from a group of clinical characteristics related to ovarian cancer.

[0016] In some embodiments, the solid tumor is a brain tumor, and the at least 2 genes selected are from the gene set described herein, wherein the gene set is capable of classifying the brain tumor as benign or malignant; and optionally wherein the one or more clinical characteristics are selected from a group of clinical characteristics related to brain cancer.

[0017] In some embodiments, the solid tumor is a kidney tumor, and the at least 2 genes selected are from the gene set described herein, wherein the gene set is capable of classifying the kidney tumor as benign or malignant; and optionally wherein the one or more clinical characteristics are selected from a group of clinical characteristics related to kidney cancer.

[0018] In some embodiments, the machine-learning model is trained according to any of the methods described herein.

[0019] In some embodiments, the method described herein, wherein: (i) the patient has cancer; (ii) the patient does not have cancer; (iii) the patient is at an elevated risk of having cancer; or (iv) the patient is asymptomatic for cancer; optionally wherein the cancer is a pancreatic cancer, an ovarian cancer, or a brain cancer.

[0020] In some embodiments, the method described herein, further comprising administering a treatment based on a solid tumor of a patient being classified as malignant.

[0021] In some embodiments, the treatment is surgery, chemotherapy, targeted therapy, immunotherapy, radiotherapy, or any combination thereof.

[0022] In some embodiments, the inference includes a confidence value between 0 and 1 that the solid tumor is malignant.

[0023] In some embodiments, the method described herein, comprising: (i) classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with an accuracy of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (ii) classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a sensitivity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iii) classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a specificity at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.; (iv) classifying the solid tumor of theAttorney Docket No.225234-718601 / PCT patient as the malignant solid tumor or the benign solid tumor with a positive predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; or (v) classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a negative predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0024] In some embodiments, a machine learning model is trained and has a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99.

[0025] Provided herein are methods for treating cancer in a patient having a solid tumor, the method comprising: (a) obtaining a data set comprising i) gene expression measurements of a biological sample from the patient, of at least 2 genes selected from the gene set described herein, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient, wherein the biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; (b) providing the data set as an input to a trained machine-learning model trained to generate an inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor; (c) receiving, as an output of the machine learning model, the inference indicating whether the data set is indicative of the malignant solid tumor or the benign solid tumor; and (d) administering a treatment based on a solid tumor of a patient being classified as a malignant tumor.

[0026] In some embodiments, the cancer is a pancreatic cancer, ovarian cancer, kidney cancer, or brain cancer.

[0027] Provided herein are systems for assessing a solid tumor of a patient, the system comprising: one or more processors; and one or more memories storing executable instructions that, as a result of execution by the one or more processors, cause the system to:Attorney Docket No.225234-718601 / PCT obtain, from a data base, a data set comprising i) gene expression measurements of a biological sample from the patient, of at least 2 genes selected from the gene set described herein, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient, wherein the biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; provide a dataset as an input to a machine-learning model trained to generate an inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor; receive, as an output of the machine learning model, the inference indicating whether a composite data set is indicative of the malignant solid tumor or the benign solid tumor; and generate a report classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor.

[0028] Provided herein are non-transitory computer-readable medium storing executable instructions for assessing a solid tumor of a patient that, as a result of execution by one or more processors of a computer system, cause the computer system to: obtain, from a data base, a data set comprising i) gene expression measurements of a biological sample from the patient, of at least 2 genes selected from the gene set described herein, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient, wherein the biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; provide a data set as an input to a machine learning model trained to generate an inference of whether the data set is indicative a malignant solid tumor or a benign solid tumor; receive, as an output of the machine learning model, the inference indicating whether a composite data set is indicative of the malignant solid tumor or the benign solid tumor; and generate a report classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor.

[0029] Provided herein are methods for obtaining a gene set capable of classifying whether a patient has cancer, the method comprising: (a) providing a dataset as an input to a machine learning classifier, said dataset comprises or is derived from gene expression measurements from a plurality of reference samples, of genes listed in a plurality of gene modules, wherein the plurality of reference samples comprises a first plurality of reference samples obtained or derived from reference subjects havingAttorney Docket No.225234-718601 / PCT cancer, and a second plurality of reference samples obtained or derived from reference subjects not having cancer; (b) performing feature selection to select a subset of gene modules from the plurality of gene modules, wherein the plurality of gene modules form features of the machine learning classifier; and (c) selecting differentially expressed genes from the genes listed in the subset of gene modules to obtain the gene set capable of classifying whether the patient has cancer.

[0030] In some embodiments, the machine learning classifier is sequential grouped feature importance (SGFI) algorithm.

[0031] In some embodiments, feature selection comprises starting from a featureless model, and sequentially adding next best feature using leave-one-group-in importance (LOGI) until no further improvement in mean misclassification error (MMCE) over an improvement threshold is achieved.

[0032] In some embodiments, the improvement threshold is 0.00001, 0.00005, 0.0001, 0.0005, or 0.001.

[0033] In some embodiments, the dataset is a batch corrected dataset.

[0034] In some embodiments, the plurality of gene modules are obtained by a method comprising: providing an initial data set comprising gene expression measurement from the plurality of reference samples, of genes of an initial gene set; selecting M genes from the initial gene set, wherein said M genes are M variably expressed genes of the initial data set, and wherein M is an integer; and clustering the M genes to obtain the plurality of gene modules.

[0035] In some embodiments, the M genes are clustered based on protein-protein interaction of proteins encoded by the M genes.

[0036] In some embodiments, the M genes are M most variably expressed genes of the initial data set.

[0037] In some embodiments, M is 500 to 10000.

[0038] In some embodiments, the method described herein, further comprising analyzing a patient data set comprising or derived from gene expression measurement of at least 2 genes selected from the genes within the gene set obtained in step (c) to classify whether a patient has cancer, wherein the gene expression measurement is obtained from a biological sample obtained or derived from the patient.

[0039] In some embodiments, the patient data set comprises or is derived from gene expression measurements of at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 110, 120, 130, 140, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172,Attorney Docket No.225234-718601 / PCT 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, or all genes selected from the genes within the gene set obtained in step (c).

[0040] In some embodiments, the patient data set comprises or is derived from gene expression measurements of the genes of the gene set obtained in step (c).

[0041] In some embodiments, the method described herein, wherein the method: (i) classifies whether the patient has cancer with an accuracy of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (ii) classifies whether the patient has cancer with a sensitivity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iii) classifies whether the patient has cancer with specificity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iv) classifies whether the patient has cancer with a positive predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; or (v) classifies whether the patient has cancer with a negative predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0042] In some embodiments, analyzing the patient data set comprises providing the patient data set as an input to a machine-learning model trained to generate an inference indicative of whether the patient has cancer based on the patient data set.

[0043] In some embodiments, the method described herein, further comprising: a) receiving as an output of the machine-learning model the inference; and b) electronically outputting a report classifying whether the patient has cancer based on the inference.

[0044] In some embodiments, the machine-learning model is trained using linear regression, logistic regression (LOG), Ridge regression, Lasso regression, elastic net (EN) regression, support vector machine (SVM), gradient boosted machine (GBM), k nearest neighbors (kNN), generalized linear model (GLM), naïve Bayes (NB) classifier, neural network, Random Forest (RF), deep learning algorithm, linear discriminant analysis (LDA), decision tree learning (DTREE), adaptive boosting (ADB), Classification and Regression Tree (CART), hierarchical clustering, or any combination thereof.Attorney Docket No.225234-718601 / PCT

[0045] In some embodiments, the machine-learning model has a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99, for classifying whether the patient has cancer.

[0046] In some embodiments, the biological sample comprises a blood sample, isolated peripheral blood mononuclear cells (PBMCs), a tissue biopsy sample, or any derivative thereof.

[0047] In some embodiments, the method described herein, further comprising selecting, recommending and / or administering a treatment to the patient, when the method classifies that the patient has cancer.

[0048] In some embodiments, the treatment comprises chemotherapy, radiation, surgery and / or any combination thereof.

[0049] In some embodiments, the cancer is a solid cancer; optionally wherein the solid cancer is adrenal cancer, anal cancer, bile duct cancer, bladder cancer, bone cancer, brain cancer, breast cancer, carcinoid cancer, cervical cancer, colorectal cancer, esophageal cancer, eye cancer, gallbladder cancer, gastrointestinal stromal tumor, germ cell cancer, head and neck cancer, kidney cancer, liver cancer, lung cancer, nasal cavity and paranasal sinus cancer, nasopharyngeal cancer, neuroblastoma, neuroendocrine cancer, oral cancer, oropharyngeal cancer, ovarian cancer, pancreatic cancer, pediatric cancer, penile cancer, pituitary cancer, prostate cancer, skin cancer, soft tissue cancer, spinal cord cancer, stomach cancer, testicular cancer, thymus cancer, thyroid cancer, ureteral cancer, uterine cancer, vaginal cancer, metastatic renal cell carcinoma, melanoma, carcinoma, a sarcoma, or vulvar cancer.

[0050] In some embodiments, the cancer is a blood cancer; optionally wherein the blood cancer is leukemia, non-Hodgkin lymphoma, Hodgkin lymphoma, an AIDS-related lymphoma, multiple myeloma, plasmacytoma, post-transplantation lymphoproliferative disorder, or Waldenstrom macroglobulinemia.

[0051] Provided herein are methods for classifying whether a patient has cancer, the method comprising: providing a patient data set comprising or derived from gene expression measurements of at least 2 genes selected from genes within the gene set obtained in step (c) described herein as an input to a machine learning model trained to generate an inference of whether the patient data set is indicative of the patient having cancer; receiving, as an output of the machine learning model the inference; and electronically outputting a report classifying whether the patient has cancer based on the inference, wherein the gene expression measurements are obtained from a biological sample obtained or derived from the patient.

[0052] In some embodiments, the patient data set comprises or is derived from gene expression measurements of at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53,Attorney Docket No.225234-718601 / PCT 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200 or all genes selected from the genes within the gene set obtained in step (c) described herein.

[0053] In some embodiments, the patient data set comprises or is derived from gene expression measurements of the genes of the gene set obtained in step (c).

[0054] In some embodiments, the patient data set is derived from the gene expression measurements using gene set variation analysis (GSVA), gene set enrichment analysis (GSEA), enrichment algorithm, multiscale embedded gene co-expression network analysis (MEGENA), weighted gene co-expression network analysis (WGCNA), differential expression analysis, Z-score, log2 expression analysis, or any combination thereof.

[0055] In some embodiments, the patient data set is derived from the gene expression measurements using GSVA.

[0056] In some embodiments, the machine learning model is trained using linear regression, logistic regression (LOG), Ridge regression, Lasso regression, elastic net (EN) regression, support vector machine (SVM), gradient boosted machine (GBM), k nearest neighbors (kNN), generalized linear model (GLM), naïve Bayes (NB) classifier, neural network, Random Forest (RF), deep learning algorithm, linear discriminant analysis (LDA), decision tree learning (DTREE), adaptive boosting (ADB), Classification and Regression Tree (CART), hierarchical clustering, or any combination thereof.

[0057] In some embodiments, the method classifies whether the patient has cancer: (i) with an accuracy of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (ii) with a sensitivity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iii) with specificity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iv) with a positive predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (v) with a negative predictive value of at least about 80%, at least about 85%, at least about 90%, at least aboutAttorney Docket No.225234-718601 / PCT 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%..

[0058] In some embodiments, the machine learning model has a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99, for classifying whether the patient has cancer.

[0059] In some embodiments, the biological sample comprises a blood sample, isolated peripheral blood mononuclear cells (PBMCs), a tissue biopsy sample, or any derivative thereof.

[0060] In some embodiments, the cancer is a solid cancer; optionally wherein the solid cancer is adrenal cancer, anal cancer, bile duct cancer, bladder cancer, bone cancer, brain cancer, breast cancer, carcinoid cancer, cervical cancer, colorectal cancer, esophageal cancer, eye cancer, gallbladder cancer, gastrointestinal stromal tumor, germ cell cancer, head and neck cancer, kidney cancer, liver cancer, lung cancer, nasal cavity and paranasal sinus cancer, nasopharyngeal cancer, neuroblastoma, neuroendocrine cancer, oral cancer, oropharyngeal cancer, ovarian cancer, pancreatic cancer, pediatric cancer, penile cancer, pituitary cancer, prostate cancer, skin cancer, soft tissue cancer, spinal cord cancer, stomach cancer, testicular cancer, thymus cancer, thyroid cancer, ureteral cancer, uterine cancer, vaginal cancer, metastatic renal cell carcinoma, melanoma, carcinoma, a sarcoma, or vulvar cancer.

[0061] In some embodiments, the cancer is a blood cancer; optionally wherein the blood cancer is leukemia, non-Hodgkin lymphoma, Hodgkin lymphoma, an AIDS-related lymphoma, multiple myeloma, plasmacytoma, post-transplantation lymphoproliferative disorder, or Waldenstrom macroglobulinemia.

[0062] In some embodiments, the method described herein, further comprising selecting, recommending and / or administering a treatment to the patient, when the method classifies that the patient has cancer.

[0063] In some embodiments, the treatment comprises chemotherapy, radiation, surgery and / or any combination thereof. INCORPORATION BY REFERENCE

[0064] All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference. To the extent publications and patents or patent applications incorporated by reference contradict the disclosure contained in the specification, the specification is intended to supersede and / or take precedence over any such contradictory material. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The novel features of the disclosure are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present disclosure will be obtained byAttorney Docket No.225234-718601 / PCT reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the disclosure are utilized, and the accompanying drawings of which:

[0066] FIG.1A is a receiver operating characteristic (ROC) plot showing performance of eight machine learning classifiers using a set of 1,178 gene features generated from ribonucleic acid (RNA) sequencing (RNA-Seq) data to distinguish malignant lung nodules versus benign lung nodules. The 1,178 genes were differentially expressed in blood samples of patients with malignant lung nodules versus benign lung nodules. The eight machine learning classifiers include LOG, GLM, kNN, RF, SVM, GBM, NB, and EN.

[0067] FIG.1B shows results of exemplary trained machine learning classifier algorithms to analyze RNA Seq data using the set of 1,178 gene features to distinguish malignant lung nodules versus benign lung nodules.

[0068] FIG.2A is a ROC plot for an optimization of differentially expressed genes to distinguish malignant lung nodules versus benign lung nodules based on an analysis of RNA-Seq data. The six machine learning classifiers include LOG, GLM, kNN, RF, SVM, and GBM.

[0069] FIG.2B shows results of exemplary trained machine learning classifier algorithms in the FIG. 2A optimization of differentially expressed genes to distinguish malignant lung nodules versus benign lung nodules.

[0070] FIG.3A is a ROC plot showing performance of eight machine learning classifiers using a set of 182 gene features generated from RNA-Seq data to distinguish malignant lung nodules versus benign lung nodules. The eight machine learning classifiers include LOG, GLM, kNN, RF, SVM, GBM, NB, and EN.

[0071] FIG.3B shows results of exemplary trained machine learning classifier algorithms to analyze RNASeq data using the set of 182 gene features to distinguish malignant lung nodules versus benign lung nodules.

[0072] FIG.4A is a ROC plot showing performance of machine learning classifiers using a set of 182 gene features generated from RNA-Seq data to distinguish malignant lung nodules versus benign lung nodules. The eight machine learning classifiers include LOG, GLM, kNN, RF, SVM, GBM, NB, and EN.

[0073] FIG.4B shows tabulated results of exemplary trained machine learning classifier algorithms corresponding to FIG.4A.

[0074] FIG.5A is a ROC plot showing performance of eight machine learning classifiers using a set of 175 gene features generated from RNA-Seq data to distinguish malignant lung nodules versus benign lung nodules. The eight machine learning classifiers include LOG, GLM, kNN, RF, SVM, GBM, NB, and EN.

[0075] FIG.5B shows tabulated results of exemplary trained machine learning classifier algorithms corresponding to FIG.5A.Attorney Docket No.225234-718601 / PCT

[0076] FIG.6A is a ROC plot showing performance of machine learning classifiers using a set of 62 gene features generated from RNA-Seq data to distinguish malignant lung nodules versus benign lung nodules. The eight machine learning classifiers include LOG, GLM, kNN, RF, SVM, GBM, NB, and EN.

[0077] FIG.6B shows tabulated results of exemplary trained machine learning classifier algorithms corresponding to FIG.6A.

[0078] FIG.7A is a ROC plot showing performance of machine learning classifiers using a set of 295 gene features generated from RNA-Seq data to distinguish malignant lung nodules versus benign lung nodules. The eight machine learning classifiers include LOG, GLM, kNN, RF, SVM, GBM, NB, and EN.

[0079] FIG.7B shows tabulated results of exemplary trained machine learning classifier algorithms corresponding to FIG.7A.

[0080] FIG.8A is a ROC plot showing performance of machine learning classifiers using a set of 175 gene features generated from RNA-Seq data to distinguish malignant lung nodules versus benign lung nodules. The eight machine learning classifiers include LOG, GLM, kNN, RF, SVM, GBM, NB, and EN.

[0081] FIG.8B shows tabulated results of exemplary trained machine learning classifier algorithms corresponding to FIG.8A.

[0082] FIG.9A is a cumulative fraction of lung nodules predicted by a logistic regression classifier using a set of 175 gene features.

[0083] FIG.9B is a cumulative fraction of lung nodules predicted by a gradient boosting classifier using a set of 175 gene features.

[0084] FIG.10 illustrates an overview of an example method 1000 for assessing a solid tumor of a subject.

[0085] FIG.11 shows a computer system 1101 that is programmed or otherwise configured to implement methods provided herein.

[0086] FIG.12 shows the correlation plot of the 8 clinical characteristics features listed in Table 6.

[0087] FIG.13A-E: FIG.13A shows ROC plots showing performance of the 9 machine learning classifiers using clinical characteristics data of the 8 clinical characteristics features listed in Table 6, to distinguish malignant lung nodules versus benign lung nodules (in 152 patients). FIG.13B shows Precision / Recall curve of the 9 machine learning classifiers using clinical characteristics data of the 8 clinical characteristics features (Table 6), to distinguish malignant lung nodules versus benign lung nodules. FIG.13C shows the tabulated results of the 9 machine learning classifiers corresponding to FIG.13A. FIG.13D shows feature importance of the 8 clinical characteristics features (Table 6) for the 9 machine learning classifiers. FIG.13E shows feature importance of the 8 clinical characteristics features for all the 9 classifiers. The 9 machine learning classifiers are LOG, RF, SVM, DTREE, ADB, NB, LDA, kNN, and GBM.Attorney Docket No.225234-718601 / PCT

[0088] FIG.14A-E: FIG.14A shows ROC plots showing performance of the 9 machine learning classifiers using clinical characteristics data of 4 clinical characteristics features, NCNSZE (nodule size), NCNUPYN (nodule in the upper lobe), AGE, and NCNMYN (Nodule Spiculated), to distinguish malignant lung nodules versus benign lung nodules. FIG.14B shows Precision / Recall curve of the 9 machine learning classifiers using clinical characteristics data of the 4 clinical characteristics features, to distinguish malignant lung nodules versus benign lung nodules. FIG.14C shows the tabulated results of the 9 machine learning classifiers corresponding to FIG.14A. FIG.14D shows feature importance of the 4 clinical characteristics features for the 9 machine learning classifiers. FIG.14E shows feature importance of the 4 clinical characteristics features for all the 9 classifiers. The 9 machine learning classifiers are LOG, RF, SVM, DTREE, ADB, NB, LDA, kNN, and GBM.

[0089] FIG.15A-E: FIG.15A shows ROC plots showing performance of the 9 machine learning classifiers using clinical characteristics data of 9 clinical characteristics features (8 features in Table 6 and cancer history) to distinguish malignant lung nodules versus benign lung nodules. FIG.15B shows Precision / Recall curve of the 9 machine learning classifiers using clinical characteristics data of the 9 clinical characteristics features, to distinguish malignant lung nodules versus benign lung nodules. FIG. 15C shows the tabulated results of the 9 machine learning classifiers corresponding to FIG.15A. FIG. 15D shows feature importance of the 9 clinical characteristics features for the 9 machine learning classifiers. FIG.15E shows feature importance of the 9 clinical characteristics features for all the 9 classifiers. The 9 machine learning classifiers are LOG, RF, SVM, DTREE, ADB, NB, LDA, kNN, and GBM.

[0090] FIG.16A-D: FIG.16A shows ROC plots showing performance of the 9 machine learning classifiers using gene expression data of the 142 gene features (Table 5), and a clinical characteristics data of 3 clinical features (NCNSZE (nodule size), NCNUPYN (nodule in the upper lobe), and AGE), to distinguish malignant lung nodules versus benign lung nodules. FIG.16B shows Precision / Recall curve of the 9 machine learning classifiers using gene expression data of the 142 gene features, and a clinical characteristics data of 3 clinical features, to distinguish malignant lung nodules versus benign lung nodules. FIG.16C shows the tabulated results of the 9 machine learning classifiers corresponding to FIG.16A. FIG.16D shows the tabulated results of the 9 machine learning classifiers corresponding to FIG.16A with oversampling correction applied (e.g.80 sample with benign lung nodule, and 80 samples with malignant lung nodule). The 9 machine learning classifiers are LOG, RF, SVM, DTREE, ADB, NB, LDA, kNN, and GBM.

[0091] FIG.17A-E: FIG.17A shows ROC plots showing performance of the 9 machine learning classifiers using measurement data of the 34 predictors (Table 7), to distinguish malignant lung nodules versus benign lung nodules. FIG.17B shows Precision / Recall curve of the 9 machine learning classifiers using measurement data of the 34 predictors, to distinguish malignant lung nodules versus benign lung nodules. FIG.17C shows the tabulated results of the machine learning classifiers LOG and RF corresponding to FIG.17A. FIG.17D shows the tabulated results of the 9 machine learning classifiersAttorney Docket No.225234-718601 / PCT corresponding to FIG.17A, with oversampling correction applied (e.g.80 sample with benign lung nodule, and 80 samples with malignant lung nodule). FIG.17 E shows feature importance of the 34 clinical characteristics features for all the 9 classifier. The 9 machine learning classifiers are LOG, RF, SVM, DTREE, ADB, NB, LDA, kNN, and GBM. FIG.18A-C: FIG.18A shows ROC plots showing performance of the 9 machine learning classifiers using gene expression data of the 175 gene features (Table 2), and a clinical characteristics data of 4 clinical features (NCNSZE (nodule size), NCNUPYN (nodule in the upper lobe), AGE, and NCNMYN (Nodule Spiculated), to distinguish malignant lung nodules versus benign lung nodules. FIG.18B shows Precision / Recall curve of the 9 machine learning classifiers using gene expression data of the 175 gene features, and a clinical characteristics data of 4 clinical features, to distinguish malignant lung nodules versus benign lung nodules. FIG.18C shows the tabulated results of the 9 machine learning classifiers corresponding to FIG.18A. The 9 machine learning classifiers are LOG, RF, SVM, DTREE, ADB, NB, LDA, kNN, and GBM. DETAILED DESCRIPTION Definitions

[0092] Various terms used throughout the present description may be read and understood as follows, unless the context indicates otherwise: “or” as used throughout is inclusive, as though written “and / or”; singular articles and pronouns as used throughout include their plural forms, and vice versa; similarly, gendered pronouns include their counterpart pronouns so that pronouns should not be understood as limiting anything described herein to use, implementation, performance, etc. by a single gender; “exemplary” should be understood as “illustrative” or “exemplifying” and not necessarily as “preferred” over other embodiments.

[0093] Further definitions for terms may be set out herein; these may apply to prior and subsequent instances of those terms, as will be understood from a reading of the present description.

[0094] Whenever the term “at least,” “greater than,” or “greater than or equal to” precedes the first numerical value in a series of two or more numerical values, the term “at least,” “greater than” or “greater than or equal to” applies to each of the numerical values in that series of numerical values. For example, greater than or equal to 1, 2, or 3 is equivalent to greater than or equal to 1, greater than or equal to 2, or greater than or equal to 3.

[0095] Whenever the term “no more than,” “less than,” or “less than or equal to” precedes the first numerical value in a series of two or more numerical values, the term “no more than,” “less than,” or “less than or equal to” applies to each of the numerical values in that series of numerical values. For example, less than or equal to 3, 2, or 1 is equivalent to less than or equal to 3, less than or equal to 2, or less than or equal to 1.Attorney Docket No.225234-718601 / PCT

[0096] The terms “subject,” “reference subject”, or “test subject” as used herein, generally refer to a human such as a patient. The subject may be a person (e.g., a patient) with a cancer, a benign solid tumor, or a malignant solid tumor; or a person that has been treated for a cancer, a benign solid tumor, or a malignant solid tumor; or a person that is being monitored for a cancer, a benign solid tumor, or a malignant solid tumor; or a person that is suspected of having a cancer, a benign solid tumor, or a malignant solid tumor; or a person that does not have or is not suspected of having a cancer, a benign solid tumor, or a malignant solid tumor.

[0097] The term “patient,” as used herein, generally refers to a human patient. The patient may be a person with a cancer, a benign solid tumor, or a malignant solid tumor; or a person that has been treated for a cancer, a benign solid tumor, or a malignant solid tumor; or a person that is being monitored for a cancer, a benign solid tumor, or a malignant solid tumor; or a person that is suspected of having a cancer, a benign solid tumor, or a malignant solid tumor; or a person that does not have or is not suspected of having a cancer, a benign solid tumor, or a malignant solid tumor. The cancer can be pancreatic cancer, ovarian cancer, and brain cancer, and the solid tumor can be a pancreatic tumor, ovarian tumor, and brain tumor respectively.

[0098] In some embodiments, the term “composite data set” or “composite” refers to or is associated with a data set comprising different data sets. In some embodiments, the composite data set comprises one or more data sets, each from different sources. Methods

[0099] In an aspect, the present disclosure provides a method for assessing a solid tumor of a patient. The method can include, any one of, any combination of, or all of steps a, b, c and d. Step a can include obtaining a data set containing i) gene expression measurements of a biological sample obtained or derived from the patient, of at least two genes selected from a gene set capable of classifying a solid tumor as benign or malignant, ii) optionally clinical characteristics data of one or more clinical characteristics of the patient. Step b, can include providing the data set as an input to a machine-learning model trained to generate an inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor. Step c, can include receiving, as an output of the machine-learning model, the inference indicating whether the data set is indicative of the malignant solid tumor or the benign solid tumor. Step d, can include electronically outputting a report classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor. The gene expression measurement of the biological sample can be performed using any suitable technique, such as any suitable RNA quantification technique, including but not limited to RNA-seq, Ampli-seq, and the like.

[0100] The solid tumor can be a lung tumor, pancreatic tumor, ovarian tumor, kidney, or brain tumor. In certain embodiments, the solid tumor is a lung tumor. In certain embodiments, the solid tumor is a pancreatic tumor. In certain embodiments, the solid tumor is an ovarian tumor. In certain embodiments, the solid tumor is a kidney tumor. In certain embodiments, the solid tumor is a brain tumor. In certainAttorney Docket No.225234-718601 / PCT embodiments, the solid tumor is a pancreatic tumor, and the at least two genes of the data set of step a, are selected from the gene set capable of classifying a pancreatic tumor as benign or malignant. In certain embodiments, the solid tumor is a pancreatic tumor, and the one or more clinical characteristics of the data set of step a, are selected from a group of clinical characteristics related to pancreatic cancer. In some particular embodiments, the solid tumor is a pancreatic tumor, and the at least two genes of the data set of step a, are selected from the gene set capable of classifying a pancreatic tumor as benign or malignant, and the one or more clinical characteristics of the data set of step a, are selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the solid tumor is an ovarian tumor, and the at least two genes of the data set of step a, are selected from the gene set capable of classifying an ovarian tumor as benign or malignant. In certain embodiments, the solid tumor is an ovarian tumor, and the one or more clinical characteristics of the data set of step a, are selected from a group of clinical characteristics related to ovarian cancer. In some particular embodiments, the solid tumor is an ovarian tumor, and the at least two genes of the data set of step a, are selected from the gene set capable of classifying an ovarian tumor as benign or malignant, and the one or more clinical characteristics of the data set of step a, are selected from a group of clinical characteristics related to ovarian cancer. In certain embodiments, the solid tumor is a kidney tumor, and the at least two genes of the data set of step a, are selected from the gene set capable of classifying a kidney tumor as benign or malignant. In certain embodiments, the solid tumor is a kidney tumor, and the one or more clinical characteristics of the data set of step a, are selected from a group of clinical characteristics related to kidney cancer. In some particular embodiments, the solid tumor is a kidney tumor, and the at least two genes of the data set of step a, are selected from the gene set capable of classifying a kidney tumor as benign or malignant, and the one or more clinical characteristics of the data set of step a, are selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the solid tumor is a brain tumor, and the at least two genes of the data set of step a, are selected from the gene set capable of classifying a brain tumor as benign or malignant. In certain embodiments, the solid tumor is a brain tumor, and the one or more clinical characteristics of the data set of step a, are selected from a group of clinical characteristics related to brain cancer. In some particular embodiments, the solid tumor is a brain tumor, and the at least two genes of the data set of step a, are selected from the gene set capable of classifying a brain tumor as benign or malignant, and the one or more clinical characteristics of the data set of step a, are selected from a group of clinical characteristics related to brain cancer. The gene set capable of classifying a solid tumor, e.g. pancreatic, ovarian, kidney, brain tumor, as benign or malignant, can be obtained or determined according to the methods (e.g. method of steps a’, b’, c’ and / or d’) described herein.

[0101] In some embodiments, the solid tumor is a lung tumor, and the data set of step a, contains i) gene expression measurements of the biological sample of at least two lung disease-associated genes selected from the group of genes listed in any one or more of Tables 1, 2, 3, 4, 5, 7, and 8, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient selected from the group of clinical characteristics listed in Table 6. In some embodiments, the at least two lung disease-Attorney Docket No.225234-718601 / PCT associated genes are selected from the group of genes listed in Table 4. In some embodiments, the at least two lung disease-associated genes are selected from the group of genes listed in Table 1. In some embodiments, the at least two lung disease-associated genes are selected from the group of genes listed in Table 2. In some embodiments, the at least two lung disease-associated genes are selected from the group of genes listed in Table 3. In some embodiments, the at least two lung disease-associated genes are selected from the group of genes listed in Table 5. In some embodiments, the at least two lung disease- associated genes are selected from the group of genes listed in Table 7. In some embodiments, the at least two lung disease-associated genes are selected from the group of genes listed in Table 8. In some embodiments, the at least two lung disease-associated genes, e.g. as of step a, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, or 182, or any value or range there between, genes selected from the group of genes listed in Table 1. In some embodiments, the at least two lung disease-associated genes, e.g. as of step a, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, or 175, or any value or range there between, genes selected from the group of genes listed in Table 2. In some embodiments, the at least two lung disease-associated genes, e.g. as of step a, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60 or 62, or any value or range there between, genes selected from the group of genes listed in Table 3. In some embodiments, the at least two lung disease-associated genes, e.g. as of step a, include at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, or 295, or any value or range there between, genes selected from the group of genes listed in Table 4. In some embodiments, the at least two lung disease-associated genes, e.g. as of step a, includes at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, or 142 genes selected from the group of genes listed in Table 5. In some embodiments, the at least two lung disease-associated genes, e.g. as of step a, includes at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes selected from the group of genes listed in Table 7. In some embodiments, the at least two lung disease- associated genes of step a, are selected from BCAT1, CRCP, COA4, OVCA2, POM121, HLA-DPA1, VPS37C, MGST2, RNF220, HDAC3, NFE2L1, WDR20, CNPY4, HOXB2, C6orf120, TMEM8A, ASAP1-IT2, C15orf54, CD101, FNBP1, TECR, PROK2, SLC35B3, TDRD9, CLHC1, LPL, IFITM3, OGFOD3, EIF2B3, TMEM65, and MKRN3. In some embodiments, the at least two lung disease-Attorney Docket No.225234-718601 / PCT associated genes, e.g. as of step a, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or 21 genes selected from the group of genes listed in Table 8.. In some embodiments, the at least two lung disease-associated genes of step a, are selected from BCAT1, USP32P2, CD177, QPCT, SCAF4, SNRPD3, BCL9L, THBS1, SLC22A18AS, ARCN1, DHX16, SATB1, ST6GAL1, CXCL1, TDRD9, ZNF831, MTCH1, FAM86HP, DHX8, RNF114, and DCTN4. These genes and those described herein are known to those of skill in the art, and described in the literature. Table A provides examples of Gene ID numbers for genes listed herein in the Tables, including Tables 7 and 8, as described in OMIM® - Online Mendelian Inheritance in Man (McKusick-Nathans Institute of Genetic Medicine, Johns Hopkins University School of Medicine, Baltimore, MD) and in the National Center for Biotechnology Information gene database (NCBI, U.S. National Library of Medicine 8600 Rockville Pike, Bethesda MD, 20894 USA), each incorporated by reference herein in their entirety.

[0102] Table A. Selected Genes Example Gene ID Numbers Entrez Gene ID Predictor OMIM No. (NCBI)Attorney Docket No.225234-718601 / PCT SLC35B3 610845 51000 TDRD9 617963 122402

[0103] In some embodiments, the solid tumor is a lung tumor, and the one or more clinical characteristics includes, 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics selected from the group of clinical characteristics listed in Table 6. In some embodiments, the solid tumor is a lung tumor, and the one or more clinical characteristics includes size of the tumor (e.g. lung nodule). In some embodiments, the solid tumor is a lung tumor, and the one or more clinical characteristics of the patient includes age of the patient. In some embodiments, the solid tumor is a lung tumor, and the one or more clinical characteristics includes presence of the nodule in the lung upper lobe. In some embodiments, the solidAttorney Docket No.225234-718601 / PCT tumor is a lung tumor, and the one or more clinical characteristics includes size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof. In some embodiments, the at least two lung disease-associated genes of step a, comprise the 31 genes listed in Table 7, and the one or more clinical characteristics of step a, comprise the size of the nodule, age of the patient, and the presence of the nodule in the lung upper lobe. In some embodiments, the at least two lung disease-associated genes of step a, consist of the 31 genes listed in Table 7, and the one or more clinical characteristics of step a, consist of the size of the nodule, age of the patient, and the presence of the nodule in the lung upper lobe. In some embodiments, the solid tumor is a lung tumor, and the data set of step a, contains i) gene expression measurements of the biological sample of at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 lung disease- associated genes selected from the group of genes listed in Table 7, and ii) clinical characteristics data of 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics of the patient selected from the group of clinical characteristics listed in Table 6. In some embodiments, the solid tumor is a lung tumor, and the data set of step a, contains i) gene expression measurements of the biological sample of at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 lung disease- associated genes selected from the group of genes listed in Table 7, and ii) clinical characteristics data of the clinical characteristics of the patient selected from size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof.

[0104] The biological sample can be blood sample, isolated peripheral blood mononuclear cells (PBMCs), solid tumor biopsy sample, nasal fluid, saliva, urine, stool, cerebrospinal fluid (CSF), or any derivative thereof. In some embodiments, the biological sample is a blood sample or any derivative thereof. In some embodiments, the biological sample is PBMCs or any derivative thereof. In some embodiments, the biological sample is a lung biopsy sample, or any derivative thereof. In some embodiments, the biological sample is a nasal fluid sample, or any derivative thereof. In some embodiments, the biological sample is a saliva sample, or any derivative thereof. In some embodiments, the biological sample is a urine sample, or any derivative thereof. In some embodiments, the biological sample is a stool sample, or any derivative thereof. In some embodiments, the biological sample is CSF sample, or any derivative thereof.

[0105] The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with an accuracy of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The machine learning model, e.g. of step b”, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with an accuracy of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more thanAttorney Docket No.225234-718601 / PCT about 99%. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with an accuracy of about 80 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with an accuracy of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with an accuracy of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with an accuracy of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with an accuracy of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0106] The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a sensitivity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The machine learning model, e.g. of step b”, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a sensitivity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at leastAttorney Docket No.225234-718601 / PCT about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a sensitivity of about 80 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a sensitivity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a sensitivity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a sensitivity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a sensitivity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0107] The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a specificity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The machine learning model, e.g. of step b”, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a specificity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%,Attorney Docket No.225234-718601 / PCT at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a specificity of about 80 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a specificity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a specificity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a specificity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a specificity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0108] The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a positive predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The machine learning model, e.g. of step b”, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a positive predictive value of at least about 50%, atAttorney Docket No.225234-718601 / PCT least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a positive predictive value of about 80 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a positive predictive value of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a positive predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a positive predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a positive predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0109] The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a negative predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than aboutAttorney Docket No.225234-718601 / PCT 99%. The machine learning model, e.g. of step b”, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a negative predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a negative predictive value of about 80 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a negative predictive value of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a negative predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a negative predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can classify the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a negative predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0110] The machine learning model, e.g. of step b, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a receiver operating characteristic (ROC) curve with an AUC of at least about 0.50, at least about 0.55, at least about 0.60, at least about 0.65, at least aboutAttorney Docket No.225234-718601 / PCT 0.70, at least about 0.75, at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99. The machine learning model, e.g. of step b, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a ROC curve with an AUC of about 0.8 to about 1. The machine learning model, e.g. of step b, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a ROC curve with an AUC of about 0.8 to about 0.85, about 0.8 to about 0.9, about 0.8 to about 0.92, about 0.8 to about 0.94, about 0.8 to about 0.95, about 0.8 to about 0.96, about 0.8 to about 0.97, about 0.8 to about 0.98, about 0.8 to about 0.99, about 0.8 to about 0.995, about 0.8 to about 1, about 0.85 to about 0.9, about 0.85 to about 0.92, about 0.85 to about 0.94, about 0.85 to about 0.95, about 0.85 to about 0.96, about 0.85 to about 0.97, about 0.85 to about 0.98, about 0.85 to about 0.99, about 0.85 to about 0.995, about 0.85 to about 1, about 0.9 to about 0.92, about 0.9 to about 0.94, about 0.9 to about 0.95, about 0.9 to about 0.96, about 0.9 to about 0.97, about 0.9 to about 0.98, about 0.9 to about 0.99, about 0.9 to about 0.995, about 0.9 to about 1, about 0.92 to about 0.94, about 0.92 to about 0.95, about 0.92 to about 0.96, about 0.92 to about 0.97, about 0.92 to about 0.98, about 0.92 to about 0.99, about 0.92 to about 0.995, about 0.92 to about 1, about 0.94 to about 0.95, about 0.94 to about 0.96, about 0.94 to about 0.97, about 0.94 to about 0.98, about 0.94 to about 0.99, about 0.94 to about 0.995, about 0.94 to about 1, about 0.95 to about 0.96, about 0.95 to about 0.97, about 0.95 to about 0.98, about 0.95 to about 0.99, about 0.95 to about 0.995, about 0.95 to about 1, about 0.96 to about 0.97, about 0.96 to about 0.98, about 0.96 to about 0.99, about 0.96 to about 0.995, about 0.96 to about 1, about 0.97 to about 0.98, about 0.97 to about 0.99, about 0.97 to about 0.995, about 0.97 to about 1, about 0.98 to about 0.99, about 0.98 to about 0.995, about 0.98 to about 1, about 0.99 to about 0.995, about 0.99 to about 1, or about 0.995 to about 1. The machine learning model, e.g. of step b, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a ROC curve with an AUC of about 0.8, about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, about 0.995, or about 1. The machine learning model, e.g. of step b, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a ROC curve with an AUC of at least about 0.8, about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, or about 0.995. The machine learning model, e.g. of step b, can infer whether the data set is indicative of a malignant solid tumor or a benign solid tumor with a ROC curve with an AUC of at most about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, about 0.995, or about 1.

[0111] The inference from the machine learning model can include a confidence value between 0 and 1, such as, 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 or 1, or any value or range there between, that the solid tumor is malignant. Higher confidence values may be correlated with a higher likelihood that the solid tumor is malignant. A malignant tumor may be characterized by or by having to ability to metastasize or grow invasively, which may be in contrast to benign tumor.Attorney Docket No.225234-718601 / PCT

[0112] In some embodiments, the patient has a cancer. In some embodiments, the patient does not have cancer. In some embodiments, the patient is suspected of having a cancer. In some embodiments, the patient is at an elevated risk of having a cancer. In some embodiments, the patient is asymptomatic for a cancer. Cancer can be lung cancer, pancreatic cancer, ovarian cancer, kidney cancer, or brain cancer. In some embodiments, the patient has pancreatic cancer. In some embodiments, the patient does not have pancreatic cancer. In some embodiments, the patient is suspected of having pancreatic cancer. In some embodiments, the patient is at an elevated risk of having a pancreatic cancer. In some embodiments, the patient is asymptomatic for pancreatic cancer. In some embodiments, the patient has ovarian cancer. In some embodiments, the patient does not have ovarian cancer. In some embodiments, the patient is suspected of having ovarian cancer. In some embodiments, the patient is at an elevated risk of having ovarian cancer. In some embodiments, the patient is asymptomatic for ovarian cancer. In some embodiments, the patient has kidney cancer. In some embodiments, the patient does not have kidney cancer. In some embodiments, the patient is suspected of having kidney cancer. In some embodiments, the patient is at an elevated risk of having a kidney cancer. In some embodiments, the patient is asymptomatic for kidney cancer. In some embodiments, the patient has brain cancer. In some embodiments, the patient does not have brain cancer. In some embodiments, the patient is suspected of having brain cancer. In some embodiments, the patient is at an elevated risk of having brain cancer. In some embodiments, the patient is asymptomatic for brain cancer. In some embodiments, the patient has a lung cancer. In some embodiments, the patient does not have lung cancer. In some embodiments, the patient is suspected of having a lung cancer. In some embodiments, the patient is at an elevated risk of having a lung cancer. In some embodiments, the patient is asymptomatic for a lung cancer.

[0113] In certain embodiments, the method includes optionally performing biopsy of the solid tumor of the patient based at least in part on the classification of the solid tumor of the patient as the malignant solid tumor or the benign solid tumor. In certain embodiments, the method includes optionally performing a biopsy of the solid tumor of the patient based at least in part on the classification of the solid tumor of the patient as the malignant solid tumor. In some embodiments, a biopsy is performed. In some embodiments, a biopsy is not performed. The decision to perform a biopsy may be made by one of skill in the art, based on knowledge and experience, in view of the classification of the lung nodule of the patient as the malignant lung nodule or the benign lung nodule. The decision to perform a biopsy may depend in part on the confidence value of the inference. In certain embodiments, biopsy of the solid tumor of the patient is not performed.

[0114] In some embodiments, the method further contains administering a treatment to the patient based at least in part on the classification of the solid tumor of the patient as the malignant solid tumor or the benign solid tumor. In some embodiments, the method contains administering a treatment to the patient based at least in part on the classification of the solid tumor of the patient as the malignant solid tumor. In some embodiments, the treatment is configured to treat a cancer of the patient. In some embodiments, the treatment is configured to reduce a severity of a cancer of the patient. In some embodiments, the treatment is configured to reduce a risk of having a cancer of the patient. The treatment can include oneAttorney Docket No.225234-718601 / PCT or more treatments of cancer. The cancer can be lung, pancreatic, ovarian, kidney, or brain cancer, and the treatment can be treatment for the lung, pancreatic, ovarian, or brain cancer respectively. In some embodiments, the method includes administering a treatment to the patient based at least in part on the classification of the pancreatic tumor of the patient as the malignant tumor or the benign tumor. In some embodiments, the method comprises administering a treatment of pancreatic cancer to the patient based at least in part on the classification of the pancreatic tumor of the patient as the malignant tumor. In some embodiments, the method includes administering a treatment to the patient based at least in part on the classification of the ovarian tumor of the patient as the malignant tumor or the benign tumor. In some embodiments, the method include administering a treatment of ovarian cancer to the patient based at least in part on the classification of the ovarian tumor of the patient as the malignant tumor. In some embodiments, the method includes administering a treatment to the patient based at least in part on the classification of the kidney tumor of the patient as the malignant tumor or the benign tumor. In some embodiments, the method comprises administering a treatment of kidney cancer to the patient based at least in part on the classification of the kidney tumor of the patient as the malignant tumor. In some embodiments, the method includes administering a treatment to the patient based at least in part on the classification of the brain tumor of the patient as the malignant tumor or the benign tumor. In some embodiments, the method includes administering a treatment of brain cancer to the patient based at least in part on the classification of the brain tumor of the patient as malignant tumor. In some embodiments, the method includes administering a treatment to the patient based at least in part on the classification of the lung tumor of the patient as the malignant tumor or the benign tumor. In some embodiments, the method comprises administering a treatment of lung cancer to the patient based at least in part on the classification of the lung tumor of the patient as the malignant tumor. The treatment can include one or more treatments of cancer. In some embodiments, the treatment is surgery, chemotherapy, targeted therapy, immunotherapy, radiotherapy, or any combination thereof.

[0115] The machine learning model of step b, can be a trained machine learning model capable of classifying a solid tumor of a patient as benign or malignant. The machine-learning model, e.g. of step b, can generate the inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor, by comparing the data set to a reference data set. The machine-learning model can be trained using the reference data set according to the methods described herein. In some embodiments, the reference data set can contain gene expression measurements of a plurality of reference biological samples from a plurality of reference subjects having solid tumor, of at least two genes selected from the gene set capable of classifying a solid tumor as benign or malignant, data regarding whether the solid tumors of the reference subjects are benign or malignant, and optionally clinical characteristics data of one or more clinical characteristics of the reference subjects. A first portion of the plurality of reference subjects can have benign solid tumor, and a second portion of the plurality of reference subjects can have malignant solid tumor. In some embodiments, the reference data set contains a plurality of individual reference data sets. A respective individual reference data set of the plurality of individual reference data sets can contain i) gene expression measurements of a reference biological sample from a referenceAttorney Docket No.225234-718601 / PCT subject having a reference solid tumor of at least two genes selected from the gene set capable of classifying a solid tumor as benign or malignant, ii) data regarding whether the reference solid tumor of the reference subject is benign or malignant, and iii) optionally clinical characteristics data of one or more clinical characteristics of the reference subject. The plurality of individual reference data sets can be obtained from the plurality of reference subjects. In some embodiments, different individual reference data sets are obtained from different reference subjects. In some embodiments, each of the individual reference data set contains i) gene expression measurements of a reference biological sample from one reference subject of the at least two genes selected from the gene set capable of classifying a solid tumor as benign or malignant, ii) data regarding whether the lung nodule of the one reference subject is benign or malignant, and iii) optionally clinical characteristics data of the one or more clinical characteristics of the one reference subject, and, wherein different individual reference data sets are obtained from different reference subjects. In certain embodiments, oversampling or undersampling correction can be made during training of the machine learning model. For example, if a reference data set includes a greater number of samples identified as benign and a relatively fewer number of samples identified as malignant, the malignant samples may be oversampled to produce a data set that has equal number of benign and malignant samples. The reference tumor, and solid tumor, can be of same type of tumor, such as both can be lung tumor, pancreatic tumor, both can be ovarian tumor, both can be kidney tumor, or both can be brain tumor. In certain embodiments, the solid tumor is a pancreatic tumor, the method can classify a pancreatic tumor as malignant or a benign, and the reference data set can contain gene expression measurements of at least two genes selected from the gene set capable of classifying a pancreatic tumor as benign or malignant. In certain embodiments, the solid tumor is a pancreatic tumor, the method can classify a pancreatic tumor as malignant or a benign, and the reference data set can contain clinical characteristic data of one or more clinical characteristics selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the solid tumor is a pancreatic tumor, the method can classify a pancreatic tumor as malignant or a benign, and the reference data set can contain i) gene expression measurements of at least two genes selected from the gene set capable of classifying a pancreatic tumor as benign or malignant, and ii) clinical characteristic data of one or more characteristics selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the solid tumor is ovarian tumor, the method can classify an ovarian tumor as malignant or a benign, and the reference data set can contain gene expression measurements of at least two genes selected from the gene set capable of classifying an ovarian tumor as benign or malignant. In certain embodiments, the solid tumor is an ovarian tumor, the method can classify an ovarian tumor as malignant or a benign, and the reference data set can contain clinical characteristic data of one or more clinical characteristics selected from a group of clinical characteristics related to ovarian cancer. In certain embodiments, the solid tumor is an ovarian tumor, the method can classify an ovarian tumor as malignant or a benign, and the reference data set can contain i) gene expression measurements of at least two genes selected from the gene set capable of classifying an ovarian tumor as benign or malignant, and ii) clinical characteristic data of one or more clinical characteristics selected from a group of clinical characteristics related to ovarian cancer.Attorney Docket No.225234-718601 / PCT In certain embodiments, the solid tumor is a kidney tumor, the method can classify a kidney tumor as malignant or a benign, and the reference data set can contain gene expression measurements of at least two genes selected from the gene set capable of classifying a kidney tumor as benign or malignant. In certain embodiments, the solid tumor is a kidney tumor, the method can classify a kidney tumor as malignant or a benign, and the reference data set can contain clinical characteristic data of one or more clinical characteristics selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the solid tumor is a kidney tumor, the method can classify a kidney tumor as malignant or a benign, and the reference data set can contain i) gene expression measurements of at least two genes selected from the gene set capable of classifying a kidney tumor as benign or malignant, and ii) clinical characteristic data of one or more characteristics selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the solid tumor is a brain tumor, the method can classify a brain tumor as malignant or a benign, and the reference data set can contain gene expression measurements of at least two genes selected from the gene set capable of classifying a brain tumor as benign or malignant. In certain embodiments, the solid tumor is a brain tumor, the method can classify a brain tumor as malignant or a benign, and the reference data set can contain clinical characteristic data of one or more clinical characteristics selected from a group of clinical characteristics related to brain cancer. In certain embodiments, the solid tumor is a brain tumor, the method can classify a brain tumor as malignant or a benign, and the reference data set can contain i) gene expression measurements of at least two genes selected from the gene set capable of classifying a brain tumor as benign or malignant, and ii) clinical characteristic data of one or more clinical characteristics selected from a group of clinical characteristics related to brain cancer. In some embodiments, the genes of the data set and genes of the reference data set can at least partially overlap. In some embodiments, clinical characteristics of the data set and clinical characteristics of the reference data set can at least partially overlap. The gene set capable of classifying a solid tumor, e.g. pancreatic, ovarian, kidney, brain tumor, as benign or malignant, can be obtained or determined according to the methods (e.g. method of steps a’, b’, c’ and / or d’) described herein.

[0116] In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the gene set of reference data set is the group of genes listed in any one or more of Tables 1, 2, 3, 4, 5, 7, and 8, and one or more clinical characteristics of the reference data set is selected from the group of clinical characteristics listed in Table 6. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the at least two genes of the reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, or 182, or any value or range there between, genes selected from the group of genes listed in Table 1. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the at least two genes of the reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37,Attorney Docket No.225234-718601 / PCT 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, or 175, or any value or range there between, genes selected from the group of genes listed in Table 2. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the at least two genes of the reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60 or 62, or any value or range there between, genes selected from the group of genes listed in Table 3. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the at least two genes of the reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, or 295, or any value or range there between, genes selected from the group of genes listed in Table 4. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the at least two genes of the reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, or 142, or any value or range there between, genes selected from the group of genes listed in Table 5. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the at least two genes of the reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes selected from the group of genes listed in Table 7. In some embodiments, the solid tumor is a lung tumor, the at least two genes of the reference data set comprise the 31 genes listed in Table 7, and the one or more clinical characteristics of the reference data set comprise the size of the nodule, age of the patient, and the presence of the nodule in the lung upper lobe. In some embodiments, the at least two genes of the reference data set consist of the 31 genes listed in Table 7, and the one or more clinical characteristics of the reference data set consist of the size of the nodule, age of the patient, and the presence of the nodule in the lung upper lobe. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the at least two genes of the reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or 21, genes selected from the group of genes listed in Table 8. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, and the one or more clinical characteristics of the reference data set include, 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics selected from the group of clinical characteristics listed in Table 6. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, and the one or more clinical characteristics of the reference data set includes size of the nodule. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, and the one or more clinical characteristics of the reference data setAttorney Docket No.225234-718601 / PCT includes age of the patient. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, and the one or more clinical characteristics of the reference data set includes presence of the nodule in the lung upper lobe. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, and the one or more clinical characteristics of the reference data set includes size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof. In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the at least two genes of the reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes selected from the group of genes listed in Table 7, and the one or more clinical characteristics of the reference data set includes 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics selected from the group of clinical characteristics listed in Table 6 In certain embodiments, the solid tumor is a lung tumor, the method can classify a lung tumor as malignant or a benign, the at least two genes of the reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes selected from the group of genes listed in Table 7, and the one or more clinical characteristics of the reference data set include size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof. The genes of the data set and genes of the reference data set can at least partially overlap and / or the optional clinical characteristics of the data set and optional clinical characteristics of the reference data set can at least partially overlap.

[0117] The reference biological sample can be blood sample, isolated peripheral blood mononuclear cells (PBMCs), solid tumor biopsy sample, nasal fluid, saliva, urine, stool, cerebrospinal fluid (CSF), or any derivative thereof. In some embodiments, the reference biological sample is a blood sample or any derivative thereof. In some embodiments, the reference biological sample is PBMCs or any derivative thereof. In some embodiments, the reference biological sample is a lung biopsy sample, or any derivative thereof. In some embodiments, the reference biological sample is a nasal fluid sample, or any derivative thereof. In some embodiments, the reference biological sample is a saliva sample, or any derivative thereof. In some embodiments, the reference biological sample is a urine sample, or any derivative thereof. In some embodiments, the reference biological sample is a stool sample, or any derivative thereof. In some embodiments, the reference biological sample is CSF sample, or any derivative thereof. The reference subjects can be human.

[0118] Gene expression data can be obtained by a data analysis tool selected from the group: a BIG-C™ big data analysis tool, an I-Scope™ big data analysis tool, a T-Scope™ big data analysis tool, a Cell Scan big data analysis tool, an MS (Molecular Signature) Scoring ™ analysis tool, and a Gene Set Variation Analysis (GSVA) tool (e.g., P-Scope).

[0119] In some embodiments, the trained machine learning model, e.g. of step b, is a supervised machine learning algorithm or an unsupervised machine learning algorithm. In some embodiments, the trained machine learning model is trained using linear regression, logistic regression (LOG), RidgeAttorney Docket No.225234-718601 / PCT regression, Lasso regression, elastic net (EN) regression, support vector machine (SVM), gradient boosted machine (GBM), k nearest neighbors (kNN), generalized linear model (GLM), naïve Bayes (NB), neural network, Random Forest (RF), deep learning algorithm, linear discriminant analysis (LDA), a decision tree learning (DTREE), adaptive boosting (ADB), or any combination thereof. In some embodiments, the trained machine learning model is trained using LOG. In some embodiments, the trained machine learning model is trained using Ridge regression. In some embodiments, the trained machine learning model is trained using Lasso regression. In some embodiments, the trained machine learning model is trained using GLM. In some embodiments, the trained machine learning model is trained using kNN. In some embodiments, the trained machine learning model is trained using SVM. In some embodiments, the trained machine learning model is trained using GBM. In some embodiments, the trained machine learning model is trained using RF. In some embodiments, the trained machine learning model is trained using NB. In some embodiments, the trained machine learning model is trained using the EN regression. In some embodiments, the trained machine learning model is trained using neural network. In some embodiments, the trained machine learning model is trained using deep learning algorithm. In some embodiments, the trained machine learning model is trained using LDA. In some embodiments, the trained machine learning model is trained using DTREE. In some embodiments, the trained machine learning model is trained using ADB.

[0120] In some embodiments, the method comprises determining a likelihood of the classification of the solid tumor of the patient as the malignant solid tumor or the benign solid tumor. In some embodiments, the likelihood is about 50%, about 55%, about 60%, about 65%, about 70%, about 75%, about 80%, about 85%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, or about 100%. In some embodiments, the likelihood is least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0121] In some embodiments, the method further comprises monitoring the solid tumor of the patient, wherein the monitoring comprises assessing the solid tumor of the patient at a plurality of time points. In some embodiments, a difference in the assessment of the solid tumor of the patient among the plurality of time points is indicative of one or more clinical indications selected from the group consisting of: (i) a diagnosis of the solid tumor of the patient, (ii) a prognosis of the solid tumor of the patient, and (iii) an efficacy or non-efficacy of a course of treatment for treating the solid tumor of the patient. In some embodiments, the plurality of time points comprises at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 different time points.

[0122] In an aspect, the present disclosure provides a method for determining a gene set capable of classifying a solid tumor, as benign or malignant. Gene expression measurements of one or more genesAttorney Docket No.225234-718601 / PCT of the gene set of a biological sample (e.g. blood) from a patient can be used to classify a solid tumor of the patient, as benign or malignant without performing biopsy of the solid tumor. In some embodiments, a biopsy of the solid tumor can be performed to confirm and / or follow-up the classification results obtained by using the gene expression measurement data. In some embodiments, a biopsy of the solid tumor is not performed. The method can include any one of, any combination of, or all of steps a’, b’, c’ and d’. In step a’, a reference data set can be obtained and / or provided.

[0123] The reference data set can contain i) gene expression measurements of a plurality of genes of reference biological samples from reference subjects each having at least one reference solid tumor, ii) data regarding whether the reference tumors are benign or malignant, and iii) optionally clinical characteristics data of one or more clinical characteristics of the reference subjects. The reference data set can contain a plurality of individual reference data sets. A respective individual reference data set of the plurality of individual reference data sets can contain i) gene expression measurements of the plurality of genes of a reference biological sample from a reference subject having a reference solid tumor, ii) data regarding whether the reference tumor is benign or malignant, and iii) optionally clinical characteristics data of the one or more clinical characteristics of the reference subject. The plurality of individual reference data sets can be obtained from a plurality of reference subjects. In some embodiments, different individual reference data sets can be obtained from different reference subjects. In some embodiments, each of the individual reference data set contains i) gene expression measurements of the plurality of genes of a reference biological sample from one reference subject, ii) data regarding whether the solid tumor of the one reference subject is benign or malignant, and iii) optionally clinical characteristics data of the one or more clinical characteristics of the one reference subject, wherein different individual reference data sets are obtained from different reference subjects. A first portion of the plurality of reference subjects can have benign solid tumor, and a second portion of the plurality of reference subjects can have malignant solid tumor. In step b’, a machine learning model can be trained using the reference data set to infer whether a solid tumor is benign or malignant based on at least in part on gene expression measurements of the plurality of genes, and optionally clinical characteristics data of the one or more clinical characteristics. In some embodiments, the machine learning model can be trained using a training data set containing a first portion of the reference data set, and a validation data set containing a second portion of the reference data set. In certain embodiments, oversampling or undersampling correction can be made during training of the machine learning model. For example, if a data set includes a greater number of samples identified as benign and a relatively fewer number of samples identified as malignant, the malignant samples may be oversampled to produce a data set that has equal number of benign and malignant samples. In step c’, feature importance values of the plurality of genes can be determined. In step d’, the gene set can be selected. The gene set can be selected as predictors that are used to train the machine learning model. The gene set, may be selected based at least in part on feature importance values. In some embodiments, the feature importance values of the genes of the gene set, are within top 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100,Attorney Docket No.225234-718601 / PCT 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 210, 220, 230, 240, or 250, or any value or range there between, feature importance values of the plurality of genes. In some embodiments, the feature importance of the genes of the gene set, can have accuracy, greater than 30 %, 35 %, 40 %, 45 %, 50 %, 55 %, 60 %, 65 %, 70 %, 75 %, or 80 % or 90 %. In some embodiments, the feature importance of the genes of the gene set, can have threshold importance, greater than 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80 or 90. In certain embodiments, the top 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 210, 220, 230, 240, or 250, or any value or range there between, predictors of the machine learning model include the genes of the gene set.

[0124] The solid tumor can be a lung tumor, pancreatic tumor, ovarian tumor, kidney tumor, or brain tumor. In certain embodiments, the solid tumor is a lung tumor. In certain embodiments, the solid tumor is a pancreatic tumor. In certain embodiments, the solid tumor is an ovarian tumor. In certain embodiments, the solid tumor is a kidney tumor. In certain embodiments, the solid tumor is a brain tumor. The reference tumor, and solid tumor, can be of same type of tumor, such as both can be lung tumor, pancreatic tumor, both can be ovarian tumor, both can be kidney tumor, or both can be brain tumor. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2 genes selected from the group of genes listed in Table 9. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2 genes selected from the group of genes listed in Table 1. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2 genes selected from the group of genes listed in Table 2. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2 genes selected from the group of genes listed in Table 3. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2 genes selected from the group of genes listed in Table 4. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2 genes selected from the group of genes listed in Table 5. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, 300, 400, 500, 600, 700, 800, 900, 1000, 1100 or 1178, or any value or range there between, genes selected from the group of genes listed in Table 9. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100,Attorney Docket No.225234-718601 / PCT 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, or 182, or any value or range there between, genes selected from the group of genes listed in Table 1. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, or 175, or any value or range there between, genes selected from the group of genes listed in Table 2. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60 or 62, or any value or range there between, genes selected from the group of genes listed in Table 3. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, or 295, or any value or range there between, genes selected from the group of genes listed in Table 4. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the reference data set of step a’, include at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, or 142, or any value or range there between, genes selected from the group of genes listed in Table 5. In some embodiments, the solid tumor is a lung tumor, and the one or more clinical characteristics of the reference data set of step a’, include 1, 2, 3, 4, 5, 6, 7, or 8, clinical characteristics selected from the group of clinical characteristics listed in Table 6. In some embodiments, the solid tumor is a lung tumor, the plurality of genes of the reference data set of step a’, include at least 2 genes selected from a group of genes listed in any one or more of Tables 1, 2, 3, 4, 5 and 9, and the one or more clinical characteristics of the reference data set of step a’, include 1, 2, 3, 4, 5, 6, 7, or 8 clinical characteristics selected from the group of clinical characteristics listed in Table 6. In certain embodiments, the solid tumor is a pancreatic tumor, and the plurality of genes of the reference data set of step a’, contains at least 2 genes selected from a group of genes related to pancreatic cancer. In certain embodiments, the solid tumor is a pancreatic tumor, and the one or more clinical characteristics of the reference data set of step a’, are selected from a group of clinical characteristics related to pancreatic cancer. In some particular embodiments, the solid tumor is a pancreatic tumor, and the plurality of genes of the reference data set of step a’, contains at least 2 genes selected from a group of genes related to pancreatic cancer, and the one or more clinical characteristics of the reference data set of step a’, are selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the solid tumor is an ovarian tumor, and the plurality of genes of the reference data set of step a’, contains at least 2 genes selected from a group of genes related to ovarian cancer. In certain embodiments, the solidAttorney Docket No.225234-718601 / PCT tumor is an ovarian tumor, and the one or more clinical characteristics of the reference data set of step a’, are selected from a group of clinical characteristics related to ovarian cancer. In some particular embodiments, the solid tumor is an ovarian tumor, and the plurality of genes of the reference data set of step a’, contains at least 2 genes selected from a group of genes related to ovarian cancer, and the one or more clinical characteristics of the reference data set of step a’, are selected from a group of clinical characteristics related to ovarian cancer. In certain embodiments, the solid tumor is a kidney tumor, and the plurality of genes of the reference data set of step a’, contains at least 2 genes selected from a group of genes related to kidney cancer. In certain embodiments, the solid tumor is a kidney tumor, and the one or more clinical characteristics of the reference data set of step a’, are selected from a group of clinical characteristics related to kidney cancer. In some particular embodiments, the solid tumor is a kidney tumor, and the plurality of genes of the reference data set of step a’, contains at least 2 genes selected from a group of genes related to kidney cancer, and the one or more clinical characteristics of the reference data set of step a’, are selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the solid tumor is a brain tumor, and the plurality of genes of the reference data set of step a’, contains at least 2 genes selected from a group of genes related to brain cancer. In certain embodiments, the solid tumor is a brain tumor, and the one or more clinical characteristics of the reference data of step a’, set are selected from a group of clinical characteristics related to brain cancer. In some particular embodiments, the solid tumor is a brain tumor, and the plurality of genes of the reference data set of step a’, contains at least 2 genes selected from a group of genes related to brain cancer, and the one or more clinical characteristics of the reference data set of step a’, are selected from a group of clinical characteristics related to brain cancer. In some embodiments, genes having collinear expression with correlation coefficients (e.g. in non-limiting aspects > 0.7 to > 0.9) were removed from the reference data set. Collinear gene expression can be measured by any suitable technique, e.g. Pearson correlation coefficient. While determining feature importance values for the plurality of genes has been described, this is merely a non-limiting illustrative example of a technique that can be practiced. In various embodiments, one or more feature selection techniques are used to determine the gene set that can classify a solid tumor benign or malignant. Feature selection techniques can include least absolute shrinkage and selection operator (Lasso) regression, support vector machine (SVM), regularized trees, decision trees, memetic algorithm, random multinomial logit (RMNL), auto-encoding networks, submodular feature selection, recursive feature elimination, or any combination thereof. In some of these cases, feature importance values need not be calculated for each of the genes of the plurality of genes. The reference biological sample can be blood sample, isolated peripheral blood mononuclear cells (PBMCs), solid tumor biopsy sample, nasal fluid, saliva, urine, stool, cerebrospinal fluid (CSF), or any derivative thereof. In some embodiments, the reference biological sample is a blood sample or any derivative thereof. In some embodiments, the reference biological sample is PBMCs or any derivative thereof. In some embodiments, the reference biological sample is a lung biopsy sample, or any derivative thereof. In some embodiments, the reference biological sample is a nasal fluid sample, or any derivative thereof. In some embodiments, the reference biological sample is a saliva sample, or any derivativeAttorney Docket No.225234-718601 / PCT thereof. In some embodiments, the reference biological sample is a urine sample, or any derivative thereof. In some embodiments, the reference biological sample is a stool sample, or any derivative thereof. In some embodiments, the reference biological sample is CSF sample, or any derivative thereof. The reference subjects can be human.

[0125] The machine learning model, e.g. of step b’, can be trained using a supervised machine learning algorithm or an unsupervised machine learning algorithm. In some embodiments, the machine learning model, e.g. of step b’, is trained using linear regression, logistic regression, Ridge regression, Lasso regression, elastic net (EN) regression, support vector machine (SVM), gradient boosted machine (GBM), k nearest neighbors (kNN), generalized linear model (GLM), naïve Bayes (NB), neural network, Random Forest (RF), deep learning algorithm, linear discriminant analysis (LDA), a decision tree learning (DTREE), adaptive boosting (ADB), or any combination thereof. In some embodiments, the machine learning model is trained using logistic regression. In some embodiments, the machine learning model is trained using Ridge regression. In some embodiments, the machine learning model is trained using Lasso regression. In some embodiments, the machine learning model is trained using GLM. In some embodiments, the machine learning model is trained using kNN. In some embodiments, the machine learning model is trained using SVM. In some embodiments, the machine learning model is trained using GBM. In some embodiments, the machine learning model is trained using RF. In some embodiments, the machine learning model is trained using NB. In some embodiments, the machine learning model is trained using the EN regression. In some embodiments, the machine learning model is trained using neural network. In some embodiments, the machine learning model is trained using deep learning algorithm. In some embodiments, the machine learning model is trained using LDA. In some embodiments, the machine learning model is trained using DTREE. In some embodiments, the machine learning model is trained using ADB.

[0126] The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with an accuracy of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with an accuracy of about 80 % to about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with an accuracy of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to aboutAttorney Docket No.225234-718601 / PCT 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with an accuracy of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with an accuracy of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with an accuracy of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0127] The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a sensitivity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a sensitivity of about 80 % to about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a sensitivity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to aboutAttorney Docket No.225234-718601 / PCT 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a sensitivity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a sensitivity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a sensitivity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0128] The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a specificity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a specificity of about 80 % to about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a specificity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. TheAttorney Docket No.225234-718601 / PCT gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a specificity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a specificity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a specificity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0129] The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a positive predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a positive predictive value of about 80 % to about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a positive predictive value of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a positive predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a positive predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %,Attorney Docket No.225234-718601 / PCT about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a positive predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0130] The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a negative predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a negative predictive value of about 80 % to about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a negative predictive value of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a negative predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a negative predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The gene set can classify a solid tumor as the malignant solid tumor or the benign solid tumor with a negative predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.Attorney Docket No.225234-718601 / PCT

[0131] In another aspect, the present disclosure provides a method for developing a trained machine learning model capable of classifying a solid tumor of a patient, as benign or malignant. The method can include any one of, any combination of, or all of steps a”, b”, c”, d” and e”. Step a”, can include obtaining and / or providing a first reference data set. In some embodiments, the first reference data set can contain i) gene expression measurements of a plurality of genes of reference biological samples from reference subjects each having at least one reference solid tumor, ii) data regarding whether the reference tumors are benign or malignant, and iii) optionally clinical characteristics data of one or more clinical characteristics of the reference subjects. The first reference data set can contain a plurality of first individual reference data sets. A respective first individual reference data set of the plurality of first individual reference data sets can contain i) gene expression measurements of the plurality of genes of a reference biological sample from a reference subject having a reference solid tumor, ii) data regarding whether the reference solid tumor of the reference subject is benign or malignant, and iii) clinical characteristics data of the one or more clinical characteristics of the reference subject. The plurality of first individual reference data sets can be obtained from a plurality of reference subjects. In some embodiments, different first individual reference data sets are obtained from different reference subjects. In some embodiments, each of the first individual reference data set contains i) gene expression measurements of the plurality of genes of a reference biological sample from one reference subject, iii) data regarding whether the reference solid tumor of the one reference subject is benign or malignant and iii) optionally clinical characteristics data of the one or more clinical characteristics of the one reference subject, wherein different first individual reference data sets are obtained from different reference subjects. A first portion of the plurality of reference subjects can have benign solid tumor, and a second portion of the plurality of reference subjects can have malignant solid tumor. In step b”, a first machine learning model can be trained using the first reference data set to infer whether a solid tumor is benign or malignant, based at least in part on one or more predictors selected from the plurality of genes, and optionally the one or more clinical characteristics. The first machine learning model can be trained to infer whether the solid tumor is benign or malignant, based at least in part on the measurement data of the plurality of genes, and optionally the clinical characteristics data of the one or more clinical characteristics. In some embodiments, the first machine learning model can be trained using a training data set containing a first portion of the first reference data set, and a validation data set containing a second portion of the first reference data set. In step c’, feature importance values of one or more predictors of the first machine learning model can be determined. In step d’, A predictors of the first machine learning model based at least in part on the feature importance values can be selected, where A can be an integer from 3 to 2000, such as 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 210, 220, 230, 240, 250, 300, 400, 500, 600, 700, 800, 900, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, or 2000, or any integer value therein. In certain embodiments, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28,Attorney Docket No.225234-718601 / PCT 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 210, 220, 230, 240, or 250, or any value or range there between, predictors of the first machine learning model can be selected. In some embodiments, the A predictors can have top A feature importance values, for example, in a non-limiting aspect, A can be 10, and 10 predictors having 10 highest feature importance values can be selected. In some embodiments, the feature importance of the A predictors, can have an accuracy, greater than 30 %, 35 %, 40 %, 45 %, 50 %, 55 %, 60 %, 65 %, 70 %, 75 %, 80 % or 90%. In some embodiments, the feature importance of the A predictors, can have a threshold importance, greater than 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80 or 90. In some embodiments, the A predictors form top 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 210, 220, 230, 240, or 250, or any value or range there between, predictors of the first machine learning model. A predictors can include one or more genes, and / or optionally one or more clinical characteristics. While determining feature importance values for the one or more predictors has been described, this is merely a non-limiting illustrative example of a technique that can be practiced. In various embodiments, one or more feature selection techniques are used to determine the A predictors. Feature selection techniques can include least absolute shrinkage and selection operator (Lasso) regression, support vector machine (SVM), regularized trees, decision trees, memetic algorithm, random multinomial logit (RMNL), auto-encoding networks, submodular feature selection, recursive feature elimination, or any combination thereof. In some of these cases, in step c”, feature importance values need not be calculated for each of the predictors of first machine learning model. Step e”, can include training a second machine learning model based at least in part on a second reference data set to obtain the trained machine learning model. The second reference data set can contain i) measurement data of the A predictors of the reference subjects, and ii) data regarding whether the solid tumors of the reference subjects are benign or malignant. The second reference data set can contain a plurality of second individual reference data sets. A respective second individual reference data set of the plurality of second individual reference data sets can include i) measurement data of the A predictors of a reference subject, and ii) data regarding whether the solid tumor of the reference subject is benign or malignant. The plurality of second individual reference data sets can be obtained from the plurality of reference subjects. In some embodiments, different second individual reference data sets are obtained from different reference subjects. In some embodiments, each of the second individual reference data set contains i) measurement data of the A predictors of one reference subject, and ii) data regarding whether the solid tumor of the one reference subject is benign or malignant, wherein different second individual reference data sets are obtained from different reference subjects. Measurement data of the A predictors can include, gene expression measurements in the reference sample of the one or more genes features of the A predictors, and / or optionally clinical characteristics data of one or more clinical characteristics features of the A predictors. The trained machine learning model can infer whether a solid tumor is benign orAttorney Docket No.225234-718601 / PCT malignant, based at least in part on measurement data of the A predictors. In some embodiments, the one or more genes features of the A predictors can form the gene set capable of classifying a solid tumor, as benign or malignant. In certain embodiments, oversampling or undersampling correction can be made during training of the first and / or second machine learning model.

[0132] The solid tumor can be a lung tumor, pancreatic tumor, ovarian tumor, kidney tumor, or a brain tumor. In certain embodiments, the solid tumor is a lung tumor. In certain embodiments, the solid tumor is a pancreatic tumor. In certain embodiments, the solid tumor is an ovarian tumor. In certain embodiments, the solid tumor is a kidney tumor. In certain embodiments, the solid tumor is a brain tumor. The reference tumor, and solid tumor, can be of same type of tumor, such as both can be lung tumor, pancreatic tumor, both can be ovarian tumor, both can be kidney tumor, or both can be brain tumor. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the first reference data set include at least 2 genes selected from a group of genes listed in Table 9. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the first reference data set include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, 300, 400, 500, 600, 700, 800, 900, 1000, 1100 or 1178, or any value or range there between, genes selected from the group of genes listed in Table 9. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the first reference data set include at least 2 genes selected from the group of genes listed in Table 1. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the first reference data set include at least 2 genes selected from the group of genes listed in Table 2. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the first reference data set include at least 2 genes selected from the group of genes listed in Table 3. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the first reference data set include at least 2 genes selected from the group of genes listed in Table 4. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the first reference data set include at least 2 genes selected from the group of genes listed in Table 5. In some embodiments, the solid tumor is a lung tumor, and the one or more clinical characteristics of the first reference data set include 1, 2, 3, 4, 5, 6, 7, or 8 clinical characteristics selected from the group of clinical characteristics listed in Table 6. In some embodiments, the solid tumor is a lung tumor, and the plurality of genes of the first reference data set include at least 2 genes selected from a group of genes listed in any one or more of Tables 1, 2, 3, 4, 5 and 9, and the one or more clinical characteristics of the first reference data set include 1, 2, 3, 4, 5, 6, 7, or 8 clinical characteristics selected from the group of clinical characteristics listed in Table 6. In certain embodiments, the solid tumor is a pancreatic tumor, and the plurality of genes of the first reference data set contains at least 2 genes selected from a group of genes related to pancreatic cancer. In certain embodiments, the solid tumor is a pancreatic tumor, and the one or more clinical characteristics of the first reference data set are selected from a group of clinical characteristics related to pancreatic cancer. InAttorney Docket No.225234-718601 / PCT some particular embodiments, the solid tumor is a pancreatic tumor, and the plurality of genes of the first reference data set contains at least 2 genes selected from a group of genes related to pancreatic cancer, and the one or more clinical characteristics of the first reference data set are selected from a group of clinical characteristics related from pancreatic cancer. In certain embodiments, the solid tumor is an ovarian tumor, and the plurality of genes of the first reference data set contains at least 2 genes selected from a group of genes related to ovarian cancer. In certain embodiments, the solid tumor is an ovarian tumor, and the one or more clinical characteristics of the first reference data set are selected from a group of clinical characteristics related to ovarian cancer. In some particular embodiments, the solid tumor is an ovarian tumor, and the plurality of genes of the first reference data set contains at least 2 genes selected from a group of genes related to ovarian cancer, and the one or more clinical characteristics of the first reference data set are selected from a group of clinical characteristics related from ovarian cancer. In certain embodiments, the solid tumor is a kidney tumor, and the plurality of genes of the first reference data set contains at least 2 genes selected from a group of genes related to kidney cancer. In certain embodiments, the solid tumor is a kidney tumor, and the one or more clinical characteristics of the first reference data set are selected from a group of clinical characteristics related to kidney cancer. In some particular embodiments, the solid tumor is a kidney tumor, and the plurality of genes of the first reference data set contains at least 2 genes selected from a group of genes related to kidney cancer, and the one or more clinical characteristics of the first reference data set are selected from a group of clinical characteristics related from kidney cancer. In certain embodiments, the solid tumor is a brain tumor, and the plurality of genes of the first reference data set contains at least 2 genes selected from a group of genes related to brain cancer. In certain embodiments, the solid tumor is a brain tumor, and the one or more clinical characteristics of the first reference data set are selected from a group of clinical characteristics related to brain cancer. In some particular embodiments, the solid tumor is a brain tumor, and the plurality of genes of the first reference data set contains at least 2 genes selected from a group of genes related to brain cancer, and the one or more clinical characteristics of the first reference data set are selected from a group of clinical characteristics related from brain cancer. In some embodiments, genes having collinear expression with correlation coefficients (e.g. in non-limiting aspects > 0.7 to > 0.9) were removed from the reference data set. Collinear gene expression can be measured by any suitable technique, e.g. Pearson correlation coefficient. In some embodiments, the solid tumor is a lung tumor, and the A predictors include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, or 182, or any value or range there between, genes selected from the group of genes listed in Table 1, as predictors. In some embodiments, the solid tumor is a lung tumor, and the A predictors include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, or 175, or any value or range there between, genes selected from the group of genes listed in Table 2, as predictors. InAttorney Docket No.225234-718601 / PCT some embodiments, the solid tumor is a lung tumor, and the A predictors include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60 or 62, or any value or range there between, genes selected from the group of genes listed in Table 3, as predictors. In some embodiments, the solid tumor is a lung tumor, and the A predictors include at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, or 295, or any value or range there between, genes selected from the group of genes listed in Table 4, as predictors. In some embodiments, the solid tumor is a lung tumor, and the A predictors include at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, or 142, or any value or range there between, genes selected from the group of genes listed in Table 5, as predictors. In some embodiments, the solid tumor is a lung tumor, and the A predictors include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes of genes selected from the group listed in Table 7, as predictors. In some embodiments, the solid tumor is a lung tumor, and the A predictors can at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or 21 genes selected from the group of genes listed in Table 8, as predictors. In some embodiments, the solid tumor is a lung tumor, and the A predictors include i) at least, 2 genes selected from the group of genes listed in any one or more of Tables 1, 2, 3, 4, 5, 7, and 8and ii) optionally size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof, as predictors. In some embodiments, the solid tumor is a lung tumor, and the A predictors can include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33 or 34 predictors selected from the group listed in Table 7. In some embodiments, the solid tumor is a lung tumor, the A predictors comprise the 34 predictors listed in Table 7. In some embodiments, the solid tumor is a lung tumor, the A predictors consist the 34 predictors listed in Table 7.

[0133] The reference biological sample can be blood sample, isolated peripheral blood mononuclear cells (PBMCs), solid tumor biopsy sample, nasal fluid, saliva, urine, stool, cerebrospinal fluid (CSF), or any derivative thereof. In some embodiments, the reference biological sample is a blood sample or any derivative thereof. In some embodiments, the reference biological sample is PBMCs or any derivative thereof. In some embodiments, the reference biological sample is a lung biopsy sample, or any derivative thereof. In some embodiments, the reference biological sample is a nasal fluid sample, or any derivative thereof. In some embodiments, the reference biological sample is a saliva sample, or any derivative thereof. In some embodiments, the reference biological sample is a urine sample, or any derivative thereof. In some embodiments, the reference biological sample is a stool sample, or any derivative thereof. In some embodiments, the reference biological sample is CSF sample, or any derivative thereof.Attorney Docket No.225234-718601 / PCT

[0134] The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with an accuracy of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with an accuracy of about 80 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with an accuracy of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with an accuracy of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with an accuracy of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with an accuracy of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a sensitivity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at leastAttorney Docket No.225234-718601 / PCT about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a sensitivity of about 80 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a sensitivity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a sensitivity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a sensitivity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a sensitivity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0136] The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a specificity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor isAttorney Docket No.225234-718601 / PCT benign or malignant with a specificity of about 80 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a specificity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a specificity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a specificity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a specificity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0137] The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a positive predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a positive predictive value of about 80 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a positive predictive value of about 80 % to about 85 %, about 80 % to about 90 %, aboutAttorney Docket No.225234-718601 / PCT 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a positive predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a positive predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a positive predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0138] The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a negative predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a negative predictive value of about 80 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a negative predictive value of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % toAttorney Docket No.225234-718601 / PCT about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a negative predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a negative predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a negative predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0139] The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.50, at least about 0.55, at least about 0.60, at least about 0.65, at least about 0.70, at least about 0.75, at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a ROC curve with an AUC of about 0.8 to about 1. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a ROC curve with an AUC of about 0.8 to about 0.85, about 0.8 to about 0.9, about 0.8 to about 0.92, about 0.8 to about 0.94, about 0.8 to about 0.95, about 0.8 to about 0.96, about 0.8 to about 0.97, about 0.8 to about 0.98, about 0.8 to about 0.99, about 0.8 to about 0.995, about 0.8 to about 1, about 0.85 to about 0.9, about 0.85 to about 0.92, about 0.85 to about 0.94, about 0.85 to about 0.95, about 0.85 to about 0.96, about 0.85 to about 0.97, about 0.85 to about 0.98, about 0.85 to about 0.99, about 0.85 to about 0.995, about 0.85 to about 1, about 0.9 to about 0.92, about 0.9 to about 0.94, about 0.9 to about 0.95, about 0.9 to about 0.96, about 0.9 to about 0.97,Attorney Docket No.225234-718601 / PCT about 0.9 to about 0.98, about 0.9 to about 0.99, about 0.9 to about 0.995, about 0.9 to about 1, about 0.92 to about 0.94, about 0.92 to about 0.95, about 0.92 to about 0.96, about 0.92 to about 0.97, about 0.92 to about 0.98, about 0.92 to about 0.99, about 0.92 to about 0.995, about 0.92 to about 1, about 0.94 to about 0.95, about 0.94 to about 0.96, about 0.94 to about 0.97, about 0.94 to about 0.98, about 0.94 to about 0.99, about 0.94 to about 0.995, about 0.94 to about 1, about 0.95 to about 0.96, about 0.95 to about 0.97, about 0.95 to about 0.98, about 0.95 to about 0.99, about 0.95 to about 0.995, about 0.95 to about 1, about 0.96 to about 0.97, about 0.96 to about 0.98, about 0.96 to about 0.99, about 0.96 to about 0.995, about 0.96 to about 1, about 0.97 to about 0.98, about 0.97 to about 0.99, about 0.97 to about 0.995, about 0.97 to about 1, about 0.98 to about 0.99, about 0.98 to about 0.995, about 0.98 to about 1, about 0.99 to about 0.995, about 0.99 to about 1, or about 0.995 to about 1. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a ROC curve with an AUC of about 0.8, about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, about 0.995, or about 1. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a ROC curve with an AUC of at least about 0.8, about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, or about 0.995. The trained machine learning model, e.g. obtained in step e”, can infer whether a solid tumor is benign or malignant with a ROC curve with an AUC of at most about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, about 0.995, or about 1.

[0140] Gene expression data can be obtained by a data analysis tool selected from the group: a BIG-C™ big data analysis tool, an I-Scope™ big data analysis tool, a T-Scope™ big data analysis tool, a Cell Scan big data analysis tool, an MS (Molecular Signature) Scoring ™ analysis tool, and a Gene Set Variation Analysis (GSVA) tool (e.g., P-Scope).

[0141] In some embodiments, the trained machine learning model is a supervised machine learning algorithm or an unsupervised machine learning algorithm. In some embodiments, first and / or second machine-learning model is independently trained using linear regression, logistic regression (LOG), Ridge regression, Lasso regression, elastic net (EN) regression, support vector machine (SVM), gradient boosted machine (GBM), k nearest neighbors (kNN), generalized linear model (GLM), naïve Bayes (NB), neural network, Random Forest (RF), deep learning algorithm, linear discriminant analysis (LDA), a decision tree learning (DTREE), adaptive boosting (ADB), or any combination thereof. In some embodiments, the first and / or second machine-learning model is independently trained using LOG. In some embodiments, the first and / or second machine-learning model is independently trained using Ridge regression. In some embodiments, the first and / or second machine-learning model is independently trained using Lasso regression. In some embodiments, the first and / or second machine-learning model is independently trained using GLM. In some embodiments, the first and / or second machine-learning model is independently trained using kNN. In some embodiments, the first and / or second machine-learning model is independently trained using SVM. In some embodiments, the first and / or second machine- learning model is independently trained using GBM. In some embodiments, the first and / or secondAttorney Docket No.225234-718601 / PCT machine-learning model is independently trained using RF. In some embodiments, the first and / or second machine-learning model is independently trained using NB. In some embodiments, the first and / or second machine-learning model is independently trained using the EN regression. In some embodiments, the first and / or second machine-learning model is independently trained using neural network. In some embodiments, the first and / or second machine-learning model is independently trained using deep learning algorithm. In some embodiments, the first and / or second machine-learning model is independently trained using LDA. In some embodiments, the first and / or second machine-learning model is independently trained using DTREE. In some embodiments, the first and / or second machine-learning model is independently trained using ADB.

[0142] In an aspect, the present disclosure provides a method for treating cancer in a patient. In some embodiments, the patient has a solid tumor. The method can include, any one of, any combination of, or all of steps a”’, b”’, c”’ and d”’. Step a”’, can include obtaining a data set containing i) gene expression measurements of a biological sample from the patient, of at least two genes selected from a gene set capable of classifying a solid tumor as benign or malignant, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient. The gene expression measurements can be obtained by assaying the biological sample. Step b’”, can include providing the data set as input to a machine- learning model trained to generate an inference of whether the data set is indicative of the patient having or not having cancer. In some embodiments, the inference infer whether the data set is indicative of the solid tumor of the patient is malignant or benign. Step c”’, can include receiving, as an output of the machine-learning model, the inference indicating whether the data set is indicative of the patient having or not having cancer. In some embodiments, the inference received as an output, indicate whether the solid tumor of the patient is malignant or the benign. Step d”’, can include administering a treatment based on the determination that the patient has cancer. In some embodiments, the treatment is be administering based on the patient’s solid tumor being classified as malignant.

[0143] The cancer can be lung, pancreatic, ovarian, kidney, or brain cancer, and the solid tumor can be a pancreatic tumor, ovarian tumor, kidney tumor, or a brain tumor respectively. In certain embodiments, the cancer is lung cancer, and the solid tumor is a lung tumor. In certain embodiments, the cancer is pancreatic cancer, and the solid tumor is a pancreatic tumor. In certain embodiments, the cancer is ovarian cancer, and the solid tumor is an ovarian tumor. In certain embodiments, the cancer is kidney cancer, and the solid tumor is a kidney tumor. In certain embodiments, the cancer is brain cancer, and the solid tumor is a brain tumor. In some embodiments, the cancer is lung cancer, the gene set of reference data set is the group of genes listed in any one or more of Tables 1, 2, 3, 4, 5, 7, and 8 and optional one or more clinical characteristics of the reference data set is selected from the group of clinical characteristics listed in Table 6. In some embodiments, the cancer is lung cancer, the dataset of step a”’, contains i) gene expression measurements of the biological sample from the patient, of at least two lung disease-associated genes selected from the group of genes listed in any one or more of Tables 1, 2, 3, 4, 5, 7, and 8, and ii) optionally clinical characteristics data of one or more clinical characteristics selected from the group of clinical characteristics listed in Table 6 of the patient. In some embodiments, theAttorney Docket No.225234-718601 / PCT cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step a”’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, or 182, or any value or range there between, genes selected from the group of genes listed in Table 1. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step a”’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, or 175, or any value or range there between, genes selected from the group of genes listed in Table 2. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step a”’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60 or 62, or any value or range there between, genes selected from the group of genes listed in Table 3. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step a”’, include at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, or 295, or any value or range there between, genes selected from the group of genes listed in Table 4. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step a”’, include at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, or 142, or any value or range there between, genes selected from the group of genes listed in Table 5. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step a”’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes selected from the group of genes listed in Table.7. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step a’”, are selected from BCAT1, CRCP, COA4, OVCA2, POM121, HLA-DPA1, VPS37C, MGST2, RNF220, HDAC3, NFE2L1, WDR20, CNPY4, HOXB2, C6orf120, TMEM8A, ASAP1-IT2, C15orf54, CD101, FNBP1, TECR, PROK2, SLC35B3, TDRD9, CLHC1, LPL, IFITM3, OGFOD3, EIF2B3, TMEM65, and MKRN3. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step a”’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or 21 genes selected from the group of genes listed in Table 8. In some embodiments, the cancer is lung cancer, the at least two lung disease-associated genes of step a”’, are selected from BCAT1, USP32P2, CD177, QPCT, SCAF4, SNRPD3, BCL9L, THBS1, SLC22A18AS, ARCN1, DHX16, SATB1, ST6GAL1, CXCL1, TDRD9, ZNF831, MTCH1, FAM86HP, DHX8,Attorney Docket No.225234-718601 / PCT RNF114, and DCTN4. In some embodiments, the cancer is lung cancer, the one or more clinical characteristics of the data set of step a’”, include 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics selected from the group of clinical characteristics listed in Table 6. In some embodiments, the cancer is lung cancer, and the one or more clinical characteristics of the data set of step a’”, include size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof. In some embodiments, the cancer is lung cancer, and the data set of step a”’, contains i) gene expression measurements of at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes selected from the group of genes listed in Table 7, and ii) clinical characteristics data of 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics selected from the group of clinical characteristics listed in Table 6. In some embodiments, the cancer is lung cancer, and the data set of step a”’, contains i) gene expression measurements of at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes selected from the group of genes listed in Table 7, and ii) clinical characteristics data of size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof. In some embodiments, the cancer is lung cancer, the at least two lung disease-associated genes of step a’”, comprise the 31 genes listed in Table 7, and the one or more clinical characteristics of step a’” comprise the size of the nodule, age of the patient, and the presence of the nodule in the lung upper lobe. In some embodiments, the cancer is lung cancer, the at least two lung disease-associated genes of step a’, consist of the 31 genes listed in Table 7, and the one or more clinical characteristics of step a”” consist of the size of the nodule, age of the patient, and the presence of the nodule in the lung upper lobe.

[0144] In certain embodiments, the cancer is pancreatic cancer, and the data set of step a”’, contains gene expression measurements of at least two genes selected from the gene set capable of classifying a pancreatic tumor as benign or malignant. In certain embodiments, the cancer is pancreatic cancer, and the data set of step a”’, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the cancer is pancreatic cancer, and the data set of step a”’, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying a pancreatic tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the cancer is ovarian cancer, and the data set of step a”’, contains gene expression measurements of at least two genes selected from the gene set capable of classifying an ovarian tumor as benign or malignant. In certain embodiments, the cancer is ovarian cancer, and the data set of step a”’, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to ovarian cancer. In certain embodiments, the cancer is ovarian cancer, and the data set of step a”’, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying an ovarian tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to ovarian cancer. In certain embodiments, the cancer is kidney cancer, and the data set of step a”’, contains gene expression measurements of at least two genesAttorney Docket No.225234-718601 / PCT selected from the gene set capable of classifying a kidney tumor as benign or malignant. In certain embodiments, the cancer is kidney cancer, and the data set of step a”’, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the cancer is kidney cancer, and the data set of step a”’, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying a kidney tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the cancer is brain cancer, and the data set of step a”’, contains gene expression measurements of at least two genes selected from the gene set capable of classifying a brain tumor as benign or malignant. In certain embodiments, the cancer is brain cancer, and the data set of step a”’, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to brain cancer. In certain embodiments, the cancer is brain cancer, and the data set of step a”’, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying a brain tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to brain cancer. The gene set capable of classifying a solid tumor, e.g. pancreatic, ovarian, kidney, brain tumor, as benign or malignant, can be obtained or determined according to the methods (e.g. method of steps a’, b’, c’ and / or d’) described herein.

[0145] The inference from the machine learning model can include a confidence value between 0 and 1, such as, 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 or 1, or any value or range there between, that the tumor is malignant, where higher confidence values may be correlated with a higher likelihood that the tumor is malignant. A malignant tumor may be characterized by the ability to metastasize or grow invasively, which may be in contrast to benign nodules.

[0146] The biological sample can be blood sample, isolated peripheral blood mononuclear cells (PBMCs), solid tumor biopsy sample, nasal fluid, saliva, urine, stool, cerebrospinal fluid (CSF), or any derivative thereof. In some embodiments, the biological sample is a blood sample or any derivative thereof. In some embodiments, the biological sample is PBMCs or any derivative thereof. In some embodiments, the biological sample is a lung biopsy sample, or any derivative thereof. In some embodiments, the biological sample is a nasal fluid sample, or any derivative thereof. In some embodiments, the biological sample is a saliva sample, or any derivative thereof. In some embodiments, the biological sample is a urine sample, or any derivative thereof. In some embodiments, the biological sample is a stool sample, or any derivative thereof. In some embodiments, the biological sample is CSF sample, or any derivative thereof.

[0147] In certain embodiments, the method includes optionally performing biopsy of the solid tumor of the patient based at least in part on the classification of the solid tumor of the patient as the malignant solid tumor or the benign solid tumor. In certain embodiments, the method includes optionally performing a biopsy of the solid tumor of the patient based at least in part on the classification of theAttorney Docket No.225234-718601 / PCT solid tumor of the patient as the malignant solid tumor. The decision to perform biopsy may depend on confidence value of the inference. In certain embodiments, biopsy of the solid tumor of the patient is not performed. The machine-learning model, e.g. of step b”’, can generate the inference of whether the data set is indicative of the patient having a malignant solid tumor or a benign solid tumor, wherein the patient having malignant solid tumor may indicate the patient has cancer, and the patient having benign solid tumor may indicate the patient does not have cancer. The machine-learning model of step b”’, can be trained according to a method described herein, e.g. according to the methods training of the machine- learning model of step b.

[0148] The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with an accuracy of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with an accuracy of about 80 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with an accuracy of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with an accuracy of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancerAttorney Docket No.225234-718601 / PCT with an accuracy of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with an accuracy of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. In some embodiments, the machine learning model of step b”’, infer whether the data set is indicative of the patient having the malignant solid tumor or benign solid tumor with an accuracy of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0149] The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a sensitivity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a sensitivity of about 80 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a sensitivity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not havingAttorney Docket No.225234-718601 / PCT cancer with a sensitivity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a sensitivity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a sensitivity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. In some embodiments, the machine learning model of step b”’, infer whether the data set is indicative of the patient having the malignant solid tumor or benign solid tumor with a sensitivity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0150] The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a specificity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a specificity of about 80 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a specificity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % toAttorney Docket No.225234-718601 / PCT about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a specificity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a specificity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a specificity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. In some embodiments, the machine learning model of step b”’, infer whether the data set is indicative of the patient having the malignant solid tumor or benign solid tumor with a specificity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0151] The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a positive predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a positive predictive value of about 80 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a positive predictive value of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to aboutAttorney Docket No.225234-718601 / PCT 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a positive predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a positive predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a positive predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. In some embodiments, the machine learning model of step b”’, infer whether the data set is indicative of the patient having the malignant solid tumor or benign solid tumor with a positive predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0152] The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a negative predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a negative predictive value of about 80 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a negative predictive value of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 %Attorney Docket No.225234-718601 / PCT to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a negative predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a negative predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a negative predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. In some embodiments, the machine learning model of step b”’, infer whether the data set is indicative of the patient having the malignant solid tumor or benign solid tumor with a negative predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0153] The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a receiver operating characteristic (ROC) curve with an Area-Under- Curve (AUC) of at least about 0.50, at least about 0.55, at least about 0.60, at least about 0.65, at least about 0.70, at least about 0.75, at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of about 0.8 to about 1. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of about 0.8 to about 0.85, about 0.8 to about 0.9, about 0.8 to about 0.92, about 0.8 to about 0.94, about 0.8 to about 0.95, about 0.8 to about 0.96, about 0.8 to about 0.97, about 0.8 to about 0.98, about 0.8 to about 0.99, about 0.8 to about 0.995, about 0.8 to about 1, about 0.85 to about 0.9, about 0.85 to about 0.92, about 0.85 to about 0.94, about 0.85 to about 0.95, about 0.85 to about 0.96, about 0.85 to about 0.97, about 0.85 to about 0.98, about 0.85 to about 0.99, about 0.85 to about 0.995, aboutAttorney Docket No.225234-718601 / PCT 0.85 to about 1, about 0.9 to about 0.92, about 0.9 to about 0.94, about 0.9 to about 0.95, about 0.9 to about 0.96, about 0.9 to about 0.97, about 0.9 to about 0.98, about 0.9 to about 0.99, about 0.9 to about 0.995, about 0.9 to about 1, about 0.92 to about 0.94, about 0.92 to about 0.95, about 0.92 to about 0.96, about 0.92 to about 0.97, about 0.92 to about 0.98, about 0.92 to about 0.99, about 0.92 to about 0.995, about 0.92 to about 1, about 0.94 to about 0.95, about 0.94 to about 0.96, about 0.94 to about 0.97, about 0.94 to about 0.98, about 0.94 to about 0.99, about 0.94 to about 0.995, about 0.94 to about 1, about 0.95 to about 0.96, about 0.95 to about 0.97, about 0.95 to about 0.98, about 0.95 to about 0.99, about 0.95 to about 0.995, about 0.95 to about 1, about 0.96 to about 0.97, about 0.96 to about 0.98, about 0.96 to about 0.99, about 0.96 to about 0.995, about 0.96 to about 1, about 0.97 to about 0.98, about 0.97 to about 0.99, about 0.97 to about 0.995, about 0.97 to about 1, about 0.98 to about 0.99, about 0.98 to about 0.995, about 0.98 to about 1, about 0.99 to about 0.995, about 0.99 to about 1, or about 0.995 to about 1. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of about 0.8, about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, about 0.995, or about 1. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of at least about 0.8, about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, or about 0.995. The machine learning model of step b”’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of at most about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, about 0.995, or about 1. In some embodiments, the machine learning model of step b”’, infer whether the data set is indicative of the patient having the malignant solid tumor or benign solid tumor with a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.50, at least about 0.55, at least about 0.60, at least about 0.65, at least about 0.70, at least about 0.75, at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99.In some embodiments, the treatment is configured to treat a cancer of the patient. In some embodiments, the treatment is configured to reduce a severity of a cancer of the patient. In some embodiments, the treatment is configured to reduce a risk of having a cancer of the patient. The treatment can include one or more treatments of cancer. The cancer can be lung, pancreatic, ovarian, kidney, or brain cancer. In certain embodiments, the cancer is lung cancer. In certain embodiments, the cancer is pancreatic cancer. In certain embodiments, the cancer is ovarian cancer. In certain embodiments, the cancer is kidney cancer. In certain embodiments, the cancer is brain cancer. In some embodiments, the data set is indicative of the patient having lung cancer, and step d”’ can include administering to the patient a treatment for lung cancer. In some embodiments, the data set is indicative of the patient having pancreatic cancer, and step d”’ can include administering to the patient a treatment for pancreatic cancer. In some embodiments, the data set is indicative of the patient having ovarian cancer, and step d”’ can include administering to the patient a treatment for ovarian cancer. In some embodiments, the data set isAttorney Docket No.225234-718601 / PCT indicative of the patient having kidney cancer, and step d”’ can include administering to the patient a treatment for kidney cancer. In some embodiments, the data set is indicative of the patient having brain cancer, and step d”’ can include administering to the patient a treatment for brain cancer. In some embodiments, the treatment is surgery, chemotherapy, targeted therapy, immunotherapy, radiotherapy, or any combination thereof.

[0154] In an aspect, the present disclosure provides a method for assessing a solid tumor of a patient, for biopsy. The method can include, any one of, any combination of, or all of steps w, x, y and z. Step w, can include obtaining a data set containing i) gene expression measurements of a biological sample obtained or derived from the patient, of at least two genes selected from the gene set capable of classifying a solid tumor as benign or malignant, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient. The data set can be obtained from assaying the biological sample. Step x, can include providing the data set as an input to a machine-learning model trained to generate an inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor. Step y, can include receiving, as an output of the machine-learning model, the inference indicating whether the data set is indicative of the malignant solid tumor or the benign solid tumor. Step z, can include performing biopsy of the solid tumor based on the solid tumor being classified as the malignant solid tumor or benign tumor. In certain embodiments, step z, can include performing biopsy of the solid tumor based on the solid tumor being classified as the malignant solid tumor. The decision to perform biopsy may depend on confidence value of the inference. In certain embodiments, biopsy of the lung nodule of the patient is not performed. The gene expression measurement of the biological sample can be performed using any suitable technique, such as any suitable RNA quantification technique, including but not limited to RNA-seq, Ampli-seq, or the like.

[0155] The solid tumor can be a lung tumor, pancreatic tumor, kidney tumor, ovarian tumor, or a brain tumor. In certain embodiments, the solid tumor is a lung tumor. In certain embodiments, the solid tumor is a pancreatic tumor. In certain embodiments, the solid tumor is an ovarian tumor. In certain embodiments, the solid tumor is a kidney tumor. In certain embodiments, the solid tumor is a brain tumor

[0156] In some embodiments, the solid tumor is a lung tumor, the data set of step w, contains i) gene expression measurements of the biological sample of at least two lung disease-associated genes selected from the group of genes listed in any one or more of Tables 1, 2, 3, 4, 5, 7, and 8 and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient selected from the group of clinical characteristics listed in Table 6.

[0157] In some embodiments, the solid tumor is a lung tumor, and the at least two lung disease- associated genes of the data set of step w, includes at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, or 182, or any value or range there between, genes selected from the group of genes listed in Table 1. In some embodiments, the solid tumor is a lung tumor, and the at leastAttorney Docket No.225234-718601 / PCT two lung disease-associated genes of the data set of step w, includes at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, or 175, or any value or range there between, genes selected from the group of genes listed in Table 2. In some embodiments, the solid tumor is a lung tumor, and the at least two lung disease-associated genes of the data set of step w, includes at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60 or 62, or any value or range there between, genes selected from the group of genes listed in Table 3. In some embodiments, the solid tumor is a lung tumor, and the at least two lung disease-associated genes of the data set of step w, includes at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, or 295, or any value or range there between, genes selected from the group of genes listed in Table 4. In some embodiments, the solid tumor is a lung tumor, and the at least two lung disease-associated genes of the data set of step w, includes at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, or 142, or any value or range there between, genes selected from the group of genes listed in Table 5. In some embodiments, the solid tumor is a lung tumor, and the at least two lung disease-associated genes of the data set of step w, includes at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes selected from the group of genes listed in Table 7. In some embodiments, the solid tumor is a lung tumor, and the at least two lung disease-associated genes of step w, are selected from BCAT1, CRCP, COA4, OVCA2, POM121, HLA-DPA1, VPS37C, MGST2, RNF220, HDAC3, NFE2L1, WDR20, CNPY4, HOXB2, C6orf120, TMEM8A, ASAP1-IT2, C15orf54, CD101, FNBP1, TECR, PROK2, SLC35B3, TDRD9, CLHC1, LPL, IFITM3, OGFOD3, EIF2B3, TMEM65, and MKRN3. In some embodiments, the solid tumor is a lung tumor, and the at least two lung disease-associated genes of the data set of step w, includes at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or 21 genes selected from the group of genes listed in Table 8. In some embodiments, the solid tumor is a lung tumor, the at least two lung disease-associated genes of step w, are selected from BCAT1, USP32P2, CD177, QPCT, SCAF4, SNRPD3, BCL9L, THBS1, SLC22A18AS, ARCN1, DHX16, SATB1, ST6GAL1, CXCL1, TDRD9, ZNF831, MTCH1, FAM86HP, DHX8, RNF114, and DCTN4. In some embodiments, the solid tumor is a lung tumor, and the one or more clinical characteristics of the data set of step w, includes 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics of the patient selected from the group listed in Table 6. In some embodiments, the solid tumor is a lung tumor, and one or more clinical characteristics of the data set of step w, includes size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof. In some embodiments, the solid tumor is aAttorney Docket No.225234-718601 / PCT lung tumor, and the data set of step w, contains i) gene expression measurements of at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 lung disease- associated genes selected from the group of genes listed in Table 7, and ii) clinical characteristics data of 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics selected from the group of clinical characteristics listed in Table 6. In some embodiments, the solid tumor is a lung tumor, and the data set of step w, contains i) gene expression measurements of at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 lung disease-associated genes selected from the group of genes listed in Table 7, and ii) clinical characteristics data of clinical characteristics of the patient selected from size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof. In some embodiments, the solid tumor is a lung tumor, the at least two lung disease-associated genes of step w, comprise the 31 genes listed in Table 7, and the one or more clinical characteristics of step w, comprise the size of the nodule, age of the patient, and the presence of the nodule in the lung upper lobe. In some embodiments, the solid tumor is a lung tumor, the at least two lung disease-associated genes of step w, consist of the 31 genes listed in Table 7, and the one or more clinical characteristics of step w, consist of the size of the nodule, age of the patient, and the presence of the nodule in the lung upper lobe.

[0158] In certain embodiments, the solid tumor is a pancreatic tumor, and the data set of step w, contains gene expression measurements of at least two genes selected from the gene set capable of classifying a pancreatic tumor as benign or malignant. In certain embodiments, the solid tumor is a pancreatic tumor, and the data set of step w, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the solid tumor is a pancreatic tumor, and the data set of step w, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying a pancreatic tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the solid tumor is an ovarian tumor, and the data set of step w, contains gene expression measurements of at least two genes selected from the gene set capable of classifying an ovarian tumor as benign or malignant. In certain embodiments, the solid tumor is an ovarian tumor, and the data set of step w, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to ovarian cancer. In certain embodiments, the solid tumor is an ovarian tumor, and the data set of step w, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying an ovarian tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to ovarian cancer. In certain embodiments, the solid tumor is a kidney tumor, and the data set of step w, contains gene expression measurements of at least two genes selected from the gene set capable of classifying a kidney tumor as benign or malignant. In certain embodiments, the solid tumor is a kidney tumor, and the data set of step w, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the solid tumor is aAttorney Docket No.225234-718601 / PCT kidney tumor, and the data set of step w, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying a kidney tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the solid tumor is a brain tumor, and the data set of step w, contains gene expression measurements of at least two genes selected from the gene set capable of classifying a brain tumor as benign or malignant. In certain embodiments, the solid tumor is a brain tumor, and the data set of step w, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to brain cancer. In certain embodiments, the solid tumor is a brain tumor, and the data set of step w, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying a brain tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to brain cancer. The gene set capable of classifying a solid tumor, e.g. pancreatic, ovarian, kidney, brain tumor, as benign or malignant, can be obtained or determined according to the methods (e.g. method of steps a’, b’, c’ and / or d’) described herein.

[0159] The biological sample can be blood sample, isolated peripheral blood mononuclear cells (PBMCs), solid tumor biopsy sample, nasal fluid, saliva, urine, stool, cerebrospinal fluid (CSF), or any derivative thereof. In some embodiments, the biological sample is a blood sample or any derivative thereof. In some embodiments, the biological sample is PBMCs or any derivative thereof. In some embodiments, the biological sample is a lung biopsy sample, or any derivative thereof. In some embodiments, the biological sample is a nasal fluid sample, or any derivative thereof. In some embodiments, the biological sample is a saliva sample, or any derivative thereof. In some embodiments, the biological sample is a urine sample, or any derivative thereof. In some embodiments, the biological sample is a stool sample, or any derivative thereof. In some embodiments, the biological sample is CSF sample, or any derivative thereof.

[0160] The machine learning model of step x, can be a trained machine learning model capable of classifying a solid tumor of a patient as benign or malignant. The machine-learning model, e.g. of step x, can be trained according to the methods described herein, e.g. as of the machine learning model of step b.

[0161] Certain aspects are directed to a method for determining cancer in a patient. The method can include, any one of, any combination of, or all of steps w’, x’, y’ and z’. Step w’ can include obtaining a data set containing i) gene expression measurements of a biological sample obtained or derived from the patient, of at least two genes selected from the gene set capable of classifying a solid tumor of the patient as benign or malignant, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient. The gene expression measurements can be obtained by assaying the biological sample. Step x’ can include providing the data set as an input to a machine-learning model trained to generate an inference of whether the data set is indicative of the patient having or not having cancer. Step y’ can include receiving, as an output of the machine-learning model, the inference indicating whether the data set is indicative of the patient having or not having cancer. Step z’ canAttorney Docket No.225234-718601 / PCT include electronically outputting a report indicating the patient has, or does not have cancer. The gene expression measurement of the biological sample can be performed using any suitable technique, such as any suitable RNA quantification technique, including but not limited to RNA-seq, Ampli-seq, or the like.

[0162] The cancer can be lung, pancreatic, ovarian, kidney, or brain cancer, and the solid tumor can be a pancreatic tumor, ovarian tumor, kidney tumor, or a brain tumor respectively. In certain embodiments, the cancer is lung cancer, and the solid tumor is a lung tumor. In certain embodiments, the cancer is pancreatic cancer, and the solid tumor is a pancreatic tumor. In certain embodiments, the cancer is ovarian cancer, and the solid tumor is an ovarian tumor. In certain embodiments, the cancer is kidney cancer, and the solid tumor is a kidney tumor. In certain embodiments, the cancer is brain cancer, and the solid tumor is a brain tumor. In some embodiments, the cancer is lung cancer, and the data set of step w’, contains i) gene expression measurements of the biological sample of at least two lung disease-associated genes selected from the group of genes listed in any one or more of Tables 1, 2, 3, 4, 5, 7, and 8 and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient selected from the group of clinical characteristics listed in Table 6. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step w’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, or 182, or any value or range there between, genes selected from the group of genes listed in Table 1. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step w’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, or 175, or any value or range there between, genes selected from the group of genes listed in Table 2. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step w’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60 or 62, or any value or range there between, genes selected from the group of genes listed in Table 3. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step w’, include at least, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, or 295, or any value or range there between, genes selected from the group of genes listed in Table 4. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step w’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, or 142 genes selected fromAttorney Docket No.225234-718601 / PCT the group of genes listed in Table 5. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step w’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 genes selected from the group of genes listed in Table 7. In some embodiments, the cancer is lung cancer, the at least two lung disease- associated genes of step w’, are selected from BCAT1, CRCP, COA4, OVCA2, POM121, HLA-DPA1, VPS37C, MGST2, RNF220, HDAC3, NFE2L1, WDR20, CNPY4, HOXB2, C6orf120, TMEM8A, ASAP1-IT2, C15orf54, CD101, FNBP1, TECR, PROK2, SLC35B3, TDRD9, CLHC1, LPL, IFITM3, OGFOD3, EIF2B3, TMEM65, and MKRN3. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of the data set of step w’, include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, or 21 genes selected from the group of genes listed in Table 8. In some embodiments, the cancer is lung cancer, the at least two lung disease-associated genes of step w, are selected from BCAT1, USP32P2, CD177, QPCT, SCAF4, SNRPD3, BCL9L, THBS1, SLC22A18AS, ARCN1, DHX16, SATB1, ST6GAL1, CXCL1, TDRD9, ZNF831, MTCH1, FAM86HP, DHX8, RNF114, and DCTN4. In some embodiments, the cancer is lung cancer, and the one or more clinical characteristics of the dataset of step w’, includes, 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics selected from the group listed in Table 6. In some embodiments, the cancer is lung cancer, and the one or more clinical characteristics of the dataset of step w’, includes size of the nodule. In some embodiments, the cancer is lung cancer, and the one or more clinical characteristics of the dataset of step w’, includes age of the patient. In some embodiments, the cancer is lung cancer, and the one or more clinical characteristics of the dataset of step w’, includes presence of the nodule in the lung upper lobe. In some embodiments, the cancer is lung cancer, and the one or more clinical characteristics of the dataset of step w’, include size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof. In some embodiments, the cancer is lung cancer, and the data set of step w’, contains i) gene expression measurements of at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 lung disease associated genes selected from the group of genes listed in Table 7, and ii) clinical characteristics data of 1, 2, 3, 4, 5, 6, 7 or 8 clinical characteristics of the patient selected from the group of clinical characteristics listed in Table 6. In some embodiments, the cancer is lung cancer, and the data set of step w’, contains i) gene expression measurements of the biological sample of at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, or 31 lung disease-associated genes selected from the group of genes listed in Table 7, and ii) clinical characteristics data of the clinical characteristics of the patient selected from size of the nodule, age of the patient, presence of the nodule in the lung upper lobe, or any combination thereof. In some embodiments, the cancer is lung cancer, and the at least two lung disease- associated genes of step w’, comprise the 31 genes listed in Table 7, and the one or more clinical characteristics of step w’, comprise the size of the nodule, age of the patient, and the presence of the nodule in the lung upper lobe. In some embodiments, the cancer is lung cancer, and the at least two lung disease-associated genes of step w’, consist of the 31 genes listed in Table 7, and the one or more clinical characteristics of step w’, consist of the size of the nodule, age of the patient, and the presence of theAttorney Docket No.225234-718601 / PCT nodule in the lung upper lobe. In certain embodiments, the cancer is pancreatic cancer, and the data set of step w’, contains gene expression measurements of at least two genes selected from the gene set capable of classifying a pancreatic tumor as benign or malignant. In certain embodiments, the cancer is pancreatic cancer, and the data set of step w’, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the cancer is pancreatic cancer, and the data set of step w’, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying a pancreatic tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to pancreatic cancer. In certain embodiments, the cancer is ovarian cancer, and the data set of step w’, contains gene expression measurements of at least two genes selected from the gene set capable of classifying an ovarian tumor as benign or malignant. In certain embodiments, the cancer is ovarian cancer, and the data set of step w’, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to ovarian cancer. In certain embodiments, the cancer is ovarian cancer, and the data set of step w’, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying an ovarian tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to ovarian cancer. In certain embodiments, the cancer is kidney cancer, and the data set of step w’, contains gene expression measurements of at least two genes selected from the gene set capable of classifying a kidney tumor as benign or malignant. In certain embodiments, the cancer is kidney cancer, and the data set of step w’, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the cancer is kidney cancer, and the data set of step w’, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying a kidney tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to kidney cancer. In certain embodiments, the cancer is brain cancer, and the data set of step w’, contains gene expression measurements of at least two genes selected from the gene set capable of classifying a brain tumor as benign or malignant. In certain embodiments, the cancer is brain cancer, and the data set of step w’, contains clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to brain cancer. In certain embodiments, the cancer is brain cancer, and the data set of step w’, contains i) gene expression measurements of at least two genes selected from the gene set capable of classifying a brain tumor as benign or malignant, and ii) clinical characteristics data of one or more clinical characteristics selected from a group of clinical characteristics related to brain cancer. The gene set capable of classifying a solid tumor, e.g. pancreatic, ovarian, kidney, brain tumor, as benign or malignant, can be obtained or determined according to the methods (e.g. method of steps a’, b’, c’ and / or d’) described herein.

[0163] The biological sample can be blood sample, isolated peripheral blood mononuclear cells (PBMCs), solid tumor biopsy sample, nasal fluid, saliva, urine, stool, cerebrospinal fluid (CSF), or anyAttorney Docket No.225234-718601 / PCT derivative thereof. In some embodiments, the biological sample is a blood sample or any derivative thereof. In some embodiments, the biological sample is PBMCs or any derivative thereof. In some embodiments, the biological sample is a lung biopsy sample, or any derivative thereof. In some embodiments, the biological sample is a nasal fluid sample, or any derivative thereof. In some embodiments, the biological sample is a saliva sample, or any derivative thereof. In some embodiments, the biological sample is a urine sample, or any derivative thereof. In some embodiments, the biological sample is a stool sample, or any derivative thereof. In some embodiments, the biological sample is CSF sample, or any derivative thereof.

[0164] The method can determine whether the patient has or does not have cancer with an accuracy of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The method can determine whether the patient has or does not have cancer with an accuracy of about 80 % to about 100 %. The method can determine whether the patient has or does not have cancer with an accuracy of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can determine whether the patient has or does not have cancer with an accuracy of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can determine whether the patient has or does not have cancer with an accuracy of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can determine whether the patient has or does not have cancer with an accuracy of at mostAttorney Docket No.225234-718601 / PCT about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0165] The method can determine whether the patient has or does not have cancer with a sensitivity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The method can determine whether the patient has or does not have cancer with a sensitivity of about 80 % to about 100 %. The method can determine whether the patient has or does not have cancer with a sensitivity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can determine whether the patient has or does not have cancer with a sensitivity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can determine whether the patient has or does not have cancer with a sensitivity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can determine whether the patient has or does not have cancer with a sensitivity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0166] The method can determine whether the patient has or does not have cancer with an specificity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%,Attorney Docket No.225234-718601 / PCT at least about 98%, at least about 99%, or more than about 99%. The method can determine whether the patient has or does not have cancer with a specificity of about 80 % to about 100 %. The method can determine whether the patient has or does not have cancer with a specificity of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can determine whether the patient has or does not have cancer with a specificity of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can determine whether the patient has or does not have cancer with a specificity of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can determine whether the patient has or does not have cancer with a specificity of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0167] The method can determine whether the patient has or does not have cancer with a positive predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The method can determine whether the patient has or does not have cancer with a positive predictive value of about 80 % to about 100 %. The method can determine whether the patient has or does not have cancer with a positive predictive value of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % toAttorney Docket No.225234-718601 / PCT about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, about 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can determine whether the patient has or does not have cancer with a positive predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can determine whether the patient has or does not have cancer with a positive predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can determine whether the patient has or does not have cancer with a positive predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0168] The method can determine whether the patient has or does not have cancer with a negative predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%. The method can determine whether the patient has or does not have cancer with a negative predictive value of about 80 % to about 100 %. The method can determine whether the patient has or does not have cancer with a negative predictive value of about 80 % to about 85 %, about 80 % to about 90 %, about 80 % to about 92 %, about 80 % to about 94 %, about 80 % to about 95 %, about 80 % to about 96 %, about 80 % to about 97 %, about 80 % to about 98 %, about 80 % to about 99 %, about 80 % to about 99.5 %, about 80 % to about 100 %, about 85 % to about 90 %, about 85 % to about 92 %, about 85 % to about 94 %, about 85 % to about 95 %, about 85 % to about 96 %, about 85 % to about 97 %, about 85 % to about 98 %, about 85 % to about 99 %, about 85 % to about 99.5 %, about 85 % to about 100 %, about 90 % to about 92 %, about 90 % to about 94 %, about 90 % to about 95 %, about 90 % to about 96 %, about 90 % to about 97 %, about 90 % to about 98 %, about 90 % to about 99 %, about 90 % to about 99.5 %, aboutAttorney Docket No.225234-718601 / PCT 90 % to about 100 %, about 92 % to about 94 %, about 92 % to about 95 %, about 92 % to about 96 %, about 92 % to about 97 %, about 92 % to about 98 %, about 92 % to about 99 %, about 92 % to about 99.5 %, about 92 % to about 100 %, about 94 % to about 95 %, about 94 % to about 96 %, about 94 % to about 97 %, about 94 % to about 98 %, about 94 % to about 99 %, about 94 % to about 99.5 %, about 94 % to about 100 %, about 95 % to about 96 %, about 95 % to about 97 %, about 95 % to about 98 %, about 95 % to about 99 %, about 95 % to about 99.5 %, about 95 % to about 100 %, about 96 % to about 97 %, about 96 % to about 98 %, about 96 % to about 99 %, about 96 % to about 99.5 %, about 96 % to about 100 %, about 97 % to about 98 %, about 97 % to about 99 %, about 97 % to about 99.5 %, about 97 % to about 100 %, about 98 % to about 99 %, about 98 % to about 99.5 %, about 98 % to about 100 %, about 99 % to about 99.5 %, about 99 % to about 100 %, or about 99.5 % to about 100 %. The method can determine whether the patient has or does not have cancer with a negative predictive value of about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %. The method can determine whether the patient has or does not have cancer with a negative predictive value of at least about 80 %, about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, or about 99.5 %. The method can determine whether the patient has or does not have cancer with a negative predictive value of at most about 85 %, about 90 %, about 92 %, about 94 %, about 95 %, about 96 %, about 97 %, about 98 %, about 99 %, about 99.5 %, or about 100 %.

[0169] The machine learning model, e.g. of step x’, can infer whether the data set is indicative of the patient having or not having cancer with a receiver operating characteristic (ROC) curve with an AUC of at least about 0.50, at least about 0.55, at least about 0.60, at least about 0.65, at least about 0.70, at least about 0.75, at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99. The machine learning model, e.g. of step x’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of about 0.8 to about 1. The machine learning model, e.g. of step x’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of about 0.8 to about 0.85, about 0.8 to about 0.9, about 0.8 to about 0.92, about 0.8 to about 0.94, about 0.8 to about 0.95, about 0.8 to about 0.96, about 0.8 to about 0.97, about 0.8 to about 0.98, about 0.8 to about 0.99, about 0.8 to about 0.995, about 0.8 to about 1, about 0.85 to about 0.9, about 0.85 to about 0.92, about 0.85 to about 0.94, about 0.85 to about 0.95, about 0.85 to about 0.96, about 0.85 to about 0.97, about 0.85 to about 0.98, about 0.85 to about 0.99, about 0.85 to about 0.995, about 0.85 to about 1, about 0.9 to about 0.92, about 0.9 to about 0.94, about 0.9 to about 0.95, about 0.9 to about 0.96, about 0.9 to about 0.97, about 0.9 to about 0.98, about 0.9 to about 0.99, about 0.9 to about 0.995, about 0.9 to about 1, about 0.92 to about 0.94, about 0.92 to about 0.95, about 0.92 to about 0.96, about 0.92 to about 0.97, about 0.92 to about 0.98, about 0.92 to about 0.99, about 0.92 to about 0.995, about 0.92 to about 1, about 0.94 to about 0.95, about 0.94 to about 0.96, about 0.94 to about 0.97, about 0.94 to about 0.98, about 0.94 to about 0.99, about 0.94 to about 0.995, about 0.94 to about 1, about 0.95 to about 0.96,Attorney Docket No.225234-718601 / PCT about 0.95 to about 0.97, about 0.95 to about 0.98, about 0.95 to about 0.99, about 0.95 to about 0.995, about 0.95 to about 1, about 0.96 to about 0.97, about 0.96 to about 0.98, about 0.96 to about 0.99, about 0.96 to about 0.995, about 0.96 to about 1, about 0.97 to about 0.98, about 0.97 to about 0.99, about 0.97 to about 0.995, about 0.97 to about 1, about 0.98 to about 0.99, about 0.98 to about 0.995, about 0.98 to about 1, about 0.99 to about 0.995, about 0.99 to about 1, or about 0.995 to about 1. The machine learning model, e.g. of step x’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of about 0.8, about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, about 0.995, or about 1. The machine learning model, e.g. of step x’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of at least about 0.8, about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, or about 0.995. The machine learning model, e.g. of step x’, can infer whether the data set is indicative of the patient having or not having cancer with a ROC curve with an AUC of at most about 0.85, about 0.9, about 0.92, about 0.94, about 0.95, about 0.96, about 0.97, about 0.98, about 0.99, about 0.995, or about 1.

[0170] The inference from the machine learning model can include a confidence value between 0 and 1, such as, 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 or 1, or any value or range there between, that the patient has cancer. Higher confidence values may be correlated with a higher likelihood that the patient has is cancer.

[0171] The machine-learning model, e.g. of step x’, can generate inference of whether the data set is indicative of the patient having a malignant solid tumor or a benign solid tumor, wherein the patient having malignant solid tumor may indicate that the patient has cancer, and patient having benign solid tumor may indicate that the patient does not have lung cancer. The machine-learning model can be trained according to a method described herein, e.g. according to the methods of training of the machine learning model of step b.

[0172] In another aspect, the present disclosure provides a computer system for assessing a solid tumor of a subject, containing: a database or other suitable data storage system that is configured to store a dataset containing a) gene expression measurements of a biological sample obtained or derived from the subject, of at least two genes selected from the gene set capable of classifying a solid tumor as benign or malignant, and ii) optionally clinical characteristics data of one or more clinical characteristics of the subject; and one or more computer processors operatively coupled to the database, wherein the one or more computer processors are individually or collectively programmed to: (i) analyze the data set to classify the solid tumor of the subject as a malignant solid tumor or a benign solid tumor; (ii) electronically output a report indicative of the classification of the solid tumor of the subject as the malignant solid tumor or the benign solid tumor. Computer-implemented methods as described herein may be executed on computer systems such as those described above. For example, a computer system may comprise one or more processors and one or more memory units that collectively store computer- readable executable instructions that, as a result of execution, cause the one or more processors toAttorney Docket No.225234-718601 / PCT collectively perform the programmed steps described above. A computer system as described herein may comprise an assay device communicatively coupled to a personal computer. The data set can be a data set (e.g. of step a) described herein. The biological sample can be a biological sample described herein.

[0173] In some embodiments, the computer system further comprises an electronic display operatively coupled to the one or more computer processors, wherein the electronic display comprises a graphical user interface that is configured to display the report.

[0174] In another aspect, the present disclosure provides one or more non-transitory computer readable media collectively containing machine-executable code that, upon execution by one or more computer processors, causes the one or more computer processors to perform a method for assessing a solid tumor of a subject, the method containing: (a) obtaining a data set containing a) gene expression measurements of a biological sample obtained or derived from a subject, of at least two genes selected from the gene set capable of classifying a solid tumor as benign or malignant, and ii) optionally clinical characteristics data of one or more clinical characteristics of the subject; (b) analyzing the data set to classify the solid tumor of the subject as a malignant solid tumor or a benign solid tumor; and (c) electronically outputting a report indicative of the classification of the solid tumor of the subject as the malignant solid tumor or the benign solid tumor. The data set can be a data set (e.g. of step a) described herein. The biological sample can be a biological sample described herein.

[0175] The disclosure includes the use of any inventive method, system, or other composition described herein, including a gene set determined using the inventive methods, for diagnosing a cancer, or for determining and / or administering a treatment of a patient or subject having a cancer.

[0176] The current disclosure includes the following aspects

[0177] Aspect 1, is directed to a method for assessing a lung nodule of a subject, comprising: (a) assaying a biological sample obtained or derived from the subject to produce a data set comprising gene expression measurements of the biological sample at each of a plurality of lung disease-associated genomic loci, wherein the plurality of disease-associated genomic loci comprises at least one gene selected from the group listed in in any one or more of Tables 1, 2, 3, 4, 5, 7, and 8; (b) analyzing the data set to classify the lung nodule of the subject as a malignant lung nodule or a benign lung nodule; and (c) electronically outputting a report indicative of the classification of the lung nodule of the subject as the malignant lung nodule or the benign lung nodule.

[0178] Aspect 2 is directed to the method of aspect 1, wherein the plurality of disease-associated genomic loci comprises at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23,Attorney Docket No.225234-718601 / PCT 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, 200, 205, 210, 215, 220, 225, 230, 235, 240, 245, 250, 255, 260, 265, 270, 275, 280, 285, 290, or 295 genes selected from the group listed in Table 4.

[0179] Aspect 3 is directed to the method of aspect 1 or 2, further comprising classifying the lung nodule of the subject as the malignant lung nodule or the benign lung nodule with an accuracy of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0180] Aspect 4 is directed to the method of any one of aspects 1 to 3, further comprising classifying the lung nodule of the subject as the malignant lung nodule or the benign lung nodule with an sensitivity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0181] Aspect 5 is directed to the method of any one of aspects 1 to 4, further comprising classifying the lung nodule of the subject as the malignant lung nodule or the benign lung nodule with an specificity of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0182] Aspect 6 is directed to the method of any one of aspects 1 to 5, further comprising classifying the lung nodule of the subject as the malignant lung nodule or the benign lung nodule with a positive predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0183] Aspect 7 is directed to the method of any one of aspects 1 to 6, further comprising classifying the lung nodule of the subject as the malignant lung nodule or the benign lung nodule with a negative predictive value of at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

[0184] Aspect 8 is directed to the method of any one of aspects 1 to 7, further comprising classifying the lung nodule of the subject as the malignant lung nodule or the benign lung nodule with an Area-Under-Attorney Docket No.225234-718601 / PCT Curve (AUC) of at least about 0.50, at least about 0.55, at least about 0.60, at least about 0.65, at least about 0.70, at least about 0.75, at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99.

[0185] Aspect 9 is directed to the method of any one of aspects 1 to 8, wherein the subject has a lung cancer.

[0186] Aspect 10 is directed to the method of any one of aspects 1 to 8, wherein the subject is suspected of having a lung cancer.

[0187] Aspect 11 is directed to the method of any one of aspects 1 to 8, wherein the subject is at elevated risk of having a lung cancer.

[0188] Aspect 12 is directed to the method of any one of aspects 1 to 8, wherein the subject is asymptomatic for a lung cancer.

[0189] Aspect 13 is directed to the method of any one of aspects 1 to 12 further comprising administering a treatment to the subject based at least in part on the classification of the lung nodule of the subject as the malignant lung nodule or the benign lung nodule.

[0190] Aspect 14 is directed to the method of aspect 13, wherein the treatment is configured to treat a lung cancer of the subject.

[0191] Aspect 15 is directed to the method of aspect 13, wherein the treatment is configured to reduce a severity of a lung cancer of the subject.

[0192] Aspect 16 is directed to the method of aspect 13, wherein the treatment is configured to reduce a risk of having a lung cancer of the subject.

[0193] Aspect 17 is directed to the method of aspect 13, wherein the treatment is selected from the group consisting of: surgery, chemotherapy, targeted therapy, immunotherapy, radiotherapy, and any combination thereof.

[0194] Aspect 18 is directed to the method of aspect 1, wherein (b) comprises using a trained machine learning classifier to analyze the data set to classify the lung nodule of the subject as the malignant lung nodule or the benign lung nodule.

[0195] Aspect 19 is directed to the method of aspect 18, wherein the trained machine learning classifier is trained using gene expression data obtained by a data analysis tool selected from the group consisting of: a BIG-C™ big data analysis tool, an I-Scope™ big data analysis tool, a T-Scope™ big data analysis tool, a CellScan big data analysis tool, an MS (Molecular Signature) Scoring ™ analysis tool, and a Gene Set Variation Analysis (GSVA) tool (e.g., P-Scope).

[0196] Aspect 20 is directed to the method of aspect 18, wherein the trained machine learning classifier is selected from the group consisting of a linear regression, a logistic regression, a Ridge regression, aAttorney Docket No.225234-718601 / PCT Lasso regression, an elastic net (EN) regression, a support vector machine (SVM), a gradient boosted machine (GBM), a k nearest neighbors (kNN), a generalized linear model (GLM), a naïve Bayes (NB) classifier, a neural network, a Random Forest (RF), a deep learning algorithm, and a combination thereof.

[0197] Aspect 21 is directed to the method of aspect 20, wherein the trained machine learning classifier comprises the logistic regression.

[0198] Aspect 22 is directed to the method of aspect 20, wherein the trained machine learning classifier comprises the GLM.

[0199] Aspect 23 is directed to the method of aspect 20, wherein the trained machine learning classifier comprises the kNN.

[0200] Aspect 24 is directed to the method of aspect 20, wherein the trained machine learning classifier comprises the SVM.

[0201] Aspect 25 is directed to the method of aspect 20, wherein the trained machine learning classifier comprises the GBM.

[0202] Aspect 26 is directed to the method of aspect 20, wherein the trained machine learning classifier comprises the RF.

[0203] Aspect 27 is directed to the method of aspect 20, wherein the trained machine learning classifier comprises the NB.

[0204] Aspect 28 is directed to the method of aspect 20, wherein the trained machine learning classifier comprises the EN regression.

[0205] Aspect 29 is directed to the method of aspect 1, wherein (b) comprises comparing the data set to a reference data set.

[0206] Aspect 30 is directed to the method of aspect 29, wherein the reference data set comprises gene expression measurements of reference biological samples at each of the plurality of lung disease- associated genomic loci.

[0207] Aspect 31 is directed to the method of aspect 29, wherein the reference biological samples comprise a first plurality of biological samples obtained or derived from subjects having a malignant lung nodule and a second plurality of biological samples obtained or derived from subjects having a benign lung nodule.

[0208] Aspect 32 is directed to the method of any one of aspects 1 to 31, wherein the biological sample is selected from the group consisting of: a blood sample, isolated peripheral blood mononuclear cells (PBMCs), a lung biopsy sample, or any derivative thereof.

[0209] Aspect 33 is directed to the method of any one of aspects 1 to 32, further comprising determining a likelihood of the classification of the lung nodule of the subject as the malignant lung nodule or the benign lung nodule.Attorney Docket No.225234-718601 / PCT

[0210] Aspect 34 is directed to the method of any one of aspects 1 to 33, further comprising monitoring the lung nodule of the subject, wherein the monitoring comprises assessing the lung nodule of the subject at a plurality of time points.

[0211] Aspect 35 is directed to the method of aspect 34, wherein a difference in the assessment of the lung nodule of the subject among the plurality of time points is indicative of one or more clinical indications selected from the group consisting of: (i) a diagnosis of the lung nodule of the subject, (ii) a prognosis of the lung nodule of...

Claims

Attorney Docket No.225234-718601 / PCT CLAIMS 1. A method for determining a gene set capable of classifying a solid tumor as benign or malignant without biopsy, the method comprising: a) obtaining a reference data set comprising a plurality of individual reference data sets, wherein a respective individual reference data set of the plurality of individual reference data sets comprises i) gene expression measurements of a plurality of genes of a reference biological sample from a reference subject having a reference solid tumor, ii) data regarding whether the reference solid tumor of the reference subject is benign or malignant, and iii) optionally clinical characteristics data of one or more clinical characteristics of the reference subject, and, wherein the reference biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; b) training a machine learning model using the reference data set, wherein the machine learning model is trained to infer whether a solid tumor is benign or malignant based at least in part on gene expression measurements of the plurality of genes, and optionally clinical characteristics data of the one or more clinical characteristics; c) determining feature importance values of the plurality of genes; and d) determining a gene set based at least in part on the feature importance values.

2. The method of claim 1, wherein the solid tumor is: (i) a pancreatic tumor, an ovarian tumor, a kidney tumor, or a brain tumor; (ii) a pancreatic tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to pancreatic cancer; (iii) an ovarian tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to ovarian cancer; (iv) a brain tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to brain cancer; (v) a kidney tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to kidney cancer; (vi) a pancreatic tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to pancreatic cancer; (vii) an ovarian tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to ovarian cancer; (viii) a brain tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to brain cancer; or (ix) a kidney tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to kidney cancer.

3. A method for developing a trained machine learning model capable of classifying a solid tumor of a patient as benign or malignant, the method comprising: (a) obtaining a first reference data set comprising a plurality of first individual reference data sets, wherein a respective first individual reference data set of the plurality of first individual reference data sets comprises i) gene expression measurements of a plurality of genes of a reference biological sample from a reference subject having a reference solid tumor, ii) data regardingAttorney Docket No.225234-718601 / PCT whether the reference solid tumor is benign or malignant, and iii) optionally clinical characteristics data of one or more clinical characteristics of the reference subject, and wherein the reference biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; (b) training a first machine learning model using the first reference data set, wherein the first machine learning model is trained to infer whether a solid tumor is benign or malignant, based at least in part on one or more predictors selected from the plurality of genes, and optionally the one or more clinical characteristics; (c) determining feature importance values of the one or more predictors of the first machine learning model; (d) selecting A predictors of the first machine learning model based at least in part on the feature importance values, wherein A is an integer from 5 to 2000; and (e) training a second machine learning model based at least in part on a second reference data set comprising a plurality of second individual reference data sets, wherein a respective second individual reference data set of the plurality of second individual reference data sets comprises i) measurement data of the A predictors of the reference subject, and ii) data regarding whether the reference solid tumor of the reference subject is benign or malignant, to obtain the trained machine learning model, wherein the trained machine learning model is trained to infer whether a solid tumor is benign or malignant, based at least in part on measurement data of the A predictors.

4. The method of claim 3, wherein the solid tumor is: (i) a pancreatic tumor, an ovarian tumor, a kidney tumor, or a brain tumor; (ii) a pancreatic tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to pancreatic cancer; (iii) an ovarian tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to ovarian cancer; (iv) a brain tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to brain cancer; (v) a kidney tumor, and the plurality of genes comprises at least 2 genes selected from a group of genes related to kidney cancer; (vi) a pancreatic tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to pancreatic cancer; (vii) an ovarian tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to ovarian cancer; (viii) a brain tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to brain cancer; (ix) a kidney tumor, and the one or more clinical characteristics are selected from a group of clinical characteristics related to kidney cancer.

5. The method of claim 3 or 4, wherein the A predictors have top 5 to 200 feature importance values.

6. The method of any one of claims 3 to 5, wherein the trained machine learning model has: (i) an accuracy of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least aboutAttorney Docket No.225234-718601 / PCT 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (ii) a sensitivity at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iii) a specificity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iv) a positive predictive value at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (v) a negative predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; or (vi) a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.

99.

7. The method of any one of claims 3 to 6, wherein the first machine learning model and second machine learning model is independently trained using a linear regression, a logistic regression, a Ridge regression, a Lasso regression, an elastic net (EN) regression, a support vector machine (SVM), a gradient boosted machine (GBM), a k nearest neighbors (kNN), a generalized linear model (GLM), a naïve Bayes (NB) classifier, a neural network, a Random Forest (RF), a deep learning algorithm, a linear discriminant analysis (LDA), a decision tree learning (DTREE), an adaptive boosting (ADB), or any combination thereof.

8. A method for assessing a solid tumor of a patient, the method comprising: a) obtaining a data set comprising i) gene expression measurements of a biological sample from the patient, of at least 2 genes selected from the gene set of claim 1, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient, wherein the biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; b) providing the data set as an input to a machine-learning model trained to generate an inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor; c) receiving, as an output of the machine-learning model, the inference indicating whether the data set is indicative of the malignant solid tumor or the benign solid tumor; and d) electronically outputting a report classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor.Attorney Docket No.225234-718601 / PCT 9. The method of claim 8, wherein the solid tumor is: (i) a pancreatic tumor, an ovarian tumor, a kidney tumor, or a brain tumor.

10. The method of claim 8, wherein the solid tumor is a pancreatic tumor, and the at least 2 genes are selected from the gene set of claim 1 or 2, wherein the gene set is capable of classifying the pancreatic tumor as benign or malignant; and optionally wherein the one or more clinical characteristics are selected from a group of clinical characteristics related to pancreatic cancer.

11. The method of claim 8, wherein the solid tumor is an ovarian tumor, and the at least 2 genes are selected from the gene set of claim 1 or 2, wherein the gene set is capable of classifying the ovarian tumor as benign or malignant; and optionally wherein the one or more clinical characteristics are selected from a group of clinical characteristics related to ovarian cancer.

12. The method of claim 8, wherein the solid tumor is a brain tumor, and the at least 2 genes selected are from the gene set of claim 1 or 2, wherein the gene set is capable of classifying the brain tumor as benign or malignant; and optionally wherein the one or more clinical characteristics are selected from a group of clinical characteristics related to brain cancer.

13. The method of claim 8, wherein the solid tumor is a kidney tumor, and the at least 2 genes selected are from the gene set of claim 1 or 2, wherein the gene set is capable of classifying the kidney tumor as benign or malignant; and optionally wherein the one or more clinical characteristics are selected from a group of clinical characteristics related to kidney cancer.

14. The method of any one of claims 8 to 13, wherein the machine-learning model is trained according to the method of any one of claims 11 to 28.

15. The method of any one of claims 8 to 14, wherein: (i) the patient has cancer; (ii) the patient does not have cancer; (iii) the patient is at an elevated risk of having cancer; or (iv) the patient is asymptomatic for cancer; optionally wherein the cancer is a pancreatic cancer, an ovarian cancer, or a brain cancer.

16. The method of any one of claims 8 to 15, further comprising administering a treatment based on a solid tumor of a patient being classified as malignant.

17. The method of claim 16, wherein the treatment is surgery, chemotherapy, targeted therapy, immunotherapy, radiotherapy, or any combination thereof.

18. The method of any one of claims 8 to 17, wherein the inference includes a confidence value between 0 and 1 that the solid tumor is malignant.

19. The method of any one of claims 8 to 18, comprising: (i) classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with an accuracy of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least aboutAttorney Docket No.225234-718601 / PCT 99%, or more than about 99%; (ii) classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a sensitivity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iii) classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a specificity at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.; (iv) classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a positive predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; or (v) classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor with a negative predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

20. The method of any one of claims 8 to 19, wherein a machine learning model is trained and has a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.

99.

21. A method for treating cancer in a patient having a solid tumor, the method comprising: (a) obtaining a data set comprising i) gene expression measurements of a biological sample from the patient, of at least 2 genes selected from the gene set of claim 1 or 2, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient, wherein the biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; (b) providing the data set as an input to a trained machine-learning model trained to generate an inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor; (c) receiving, as an output of the machine learning model, the inference indicating whether the data set is indicative of the malignant solid tumor or the benign solid tumor; and (d) administering a treatment based on a solid tumor of a patient being classified as a malignant tumor.

22. The method of claim 21, wherein the cancer is a pancreatic cancer, ovarian cancer, kidney cancer, or brain cancer.Attorney Docket No.225234-718601 / PCT 23. A system for assessing a solid tumor of a patient, the system comprising: one or more processors; and one or more memories storing executable instructions that, as a result of execution by the one or more processors, cause the system to: obtain, from a data base, a data set comprising i) gene expression measurements of a biological sample from the patient, of at least 2 genes selected from the gene set of claims 1 or 2, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient, wherein the biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; provide a dataset as an input to a machine-learning model trained to generate an inference of whether the data set is indicative of a malignant solid tumor or a benign solid tumor; receive, as an output of the machine learning model, the inference indicating whether a composite data set is indicative of the malignant solid tumor or the benign solid tumor; and generate a report classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor.

24. A non-transitory computer-readable medium storing executable instructions for assessing a solid tumor of a patient that, as a result of execution by one or more processors of a computer system, cause the computer system to: obtain, from a data base, a data set comprising i) gene expression measurements of a biological sample from the patient, of at least 2 genes selected from the gene set of claims 1 or 2, and ii) optionally clinical characteristics data of one or more clinical characteristics of the patient, wherein the biological sample is a blood sample, isolated peripheral blood mononuclear cells (PBMCs), or any derivative thereof; provide a data set as an input to a machine learning model trained to generate an inference of whether the data set is indicative a malignant solid tumor or a benign solid tumor; receive, as an output of the machine learning model, the inference indicating whether a composite data set is indicative of the malignant solid tumor or the benign solid tumor; and generate a report classifying the solid tumor of the patient as the malignant solid tumor or the benign solid tumor.

25. A method for obtaining a gene set capable of classifying whether a patient has cancer, the method comprising: (a) providing a dataset as an input to a machine learning classifier, said dataset comprises or is derived from gene expression measurements from a plurality of reference samples, of genes listed in a plurality of gene modules, wherein the plurality of reference samples comprises aAttorney Docket No.225234-718601 / PCT first plurality of reference samples obtained or derived from reference subjects having cancer, and a second plurality of reference samples obtained or derived from reference subjects not having cancer; (b) performing feature selection to select a subset of gene modules from the plurality of gene modules, wherein the plurality of gene modules form features of the machine learning classifier; and (c) selecting differentially expressed genes from the genes listed in the subset of gene modules to obtain the gene set capable of classifying whether the patient has cancer.

26. The method of claim 25, wherein the machine learning classifier is sequential grouped feature importance (SGFI) algorithm.

27. The method of claim 25 or 26, wherein the feature selection comprises starting from a featureless model, and sequentially adding next best feature using leave-one-group-in importance (LOGI) until no further improvement in mean misclassification error (MMCE) over an improvement threshold is achieved.

28. The method of claim 27, wherein the improvement threshold is 0.00001, 0.00005, 0.0001, 0.0005, or 0.

001.

29. The method of any one of claims 25 to 28, wherein the dataset is a batch corrected dataset.

30. The method of any one of claims 25 to 29, wherein the plurality of gene modules are obtained by a method comprising: providing an initial data set comprising gene expression measurement from the plurality of reference samples, of genes of an initial gene set; selecting M genes from the initial gene set, wherein said M genes are M variably expressed genes of the initial data set, and wherein M is an integer; and clustering the M genes to obtain the plurality of gene modules.

31. The method of claim 30, wherein the M genes are clustered based on protein-protein interaction of proteins encoded by the M genes.

32. The method of claim 30 to 31, wherein the M genes are M most variably expressed genes of the initial data set.

33. The method of any one of claims 30 to 32, wherein M is 500 to 10000.

34. The method of any one of claims 25 to 33, further comprising analyzing a patient data set comprising or derived from gene expression measurement of at least 2 genes selected from the genes within the gene set obtained in step (c) to classify whether a patient has cancer, wherein the gene expression measurement is obtained from a biological sample obtained or derived from the patient.

35. The method of claim 34, wherein the patient data set comprises or is derived from gene expression measurements of at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72,Attorney Docket No.225234-718601 / PCT 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 110, 120, 130, 140, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, or all genes selected from the genes within the gene set obtained in step (c).

36. The method of claim 34, wherein the patient data set comprises or is derived from gene expression measurements of the genes of the gene set obtained in step (c).

37. The method of any one of claims 34 to 36, wherein the method: (i) classifies whether the patient has cancer with an accuracy of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (ii) classifies whether the patient has cancer with a sensitivity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iii) classifies whether the patient has cancer with specificity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iv) classifies whether the patient has cancer with a positive predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; or (v) classifies whether the patient has cancer with a negative predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%.

38. The method of any one of claims 34 to 37, wherein analyzing the patient data set comprises providing the patient data set as an input to a machine-learning model trained to generate an inference indicative of whether the patient has cancer based on the patient data set.

39. The method of claim 38, further comprising: a) receiving as an output of the machine-learning model the inference; and b) electronically outputting a report classifying whether the patient has cancer based on the inference.

40. The method of claim 38 or 39, wherein the machine-learning model is trained using linear regression, logistic regression (LOG), Ridge regression, Lasso regression, elastic net (EN) regression, support vector machine (SVM), gradient boosted machine (GBM), k nearest neighbors (kNN), generalized linear model (GLM), naïve Bayes (NB) classifier, neural network, Random Forest (RF), deep learning algorithm, linear discriminant analysis (LDA), decision treeAttorney Docket No.225234-718601 / PCT learning (DTREE), adaptive boosting (ADB), Classification and Regression Tree (CART), hierarchical clustering, or any combination thereof.

41. The method of any one of claims 38 to 40, wherein the machine-learning model has a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99, for classifying whether the patient has cancer.

42. The method of any one of claims 34 to 41, wherein the biological sample comprises a blood sample, isolated peripheral blood mononuclear cells (PBMCs), a tissue biopsy sample, or any derivative thereof.

43. The method of any one of claims 34 to 42, further comprising selecting, recommending and / or administering a treatment to the patient, when the method classifies that the patient has cancer.

44. The method of claim 43, wherein the treatment comprises chemotherapy, radiation, surgery and / or any combination thereof.

45. The method of any one of claims 25 to 44, wherein the cancer is a solid cancer; optionally wherein the solid cancer is adrenal cancer, anal cancer, bile duct cancer, bladder cancer, bone cancer, brain cancer, breast cancer, carcinoid cancer, cervical cancer, colorectal cancer, esophageal cancer, eye cancer, gallbladder cancer, gastrointestinal stromal tumor, germ cell cancer, head and neck cancer, kidney cancer, liver cancer, lung cancer, nasal cavity and paranasal sinus cancer, nasopharyngeal cancer, neuroblastoma, neuroendocrine cancer, oral cancer, oropharyngeal cancer, ovarian cancer, pancreatic cancer, pediatric cancer, penile cancer, pituitary cancer, prostate cancer, skin cancer, soft tissue cancer, spinal cord cancer, stomach cancer, testicular cancer, thymus cancer, thyroid cancer, ureteral cancer, uterine cancer, vaginal cancer, metastatic renal cell carcinoma, melanoma, carcinoma, a sarcoma, or vulvar cancer.

46. The method of any one of claims 25 to 44, wherein the cancer is a blood cancer; optionally wherein the blood cancer is leukemia, non-Hodgkin lymphoma, Hodgkin lymphoma, an AIDS- related lymphoma, multiple myeloma, plasmacytoma, post-transplantation lymphoproliferative disorder, or Waldenstrom macroglobulinemia.

47. A method for classifying whether a patient has cancer, the method comprising: providing a patient data set comprising or derived from gene expression measurements of at least 2 genes selected from genes within the gene set obtained in step (c) of any one of claims 25- 33 as an input to a machine learning model trained to generate an inference of whether the patient data set is indicative of the patient having cancer; receiving, as an output of the machine learning model the inference; and electronically outputting a report classifying whether the patient has cancer based on the inference,Attorney Docket No.225234-718601 / PCT wherein the gene expression measurements are obtained from a biological sample obtained or derived from the patient.

48. The method of claim 47, wherein the patient data set comprises or is derived from gene expression measurements of at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200 or all genes selected from the genes within the gene set obtained in step (c) of any one of claims 25-33.

49. The method of claim 47, wherein the patient data set comprises or is derived from gene expression measurements of the genes of the gene set obtained in step (c).

50. The method of any one of claims 47 to 49, wherein the patient data set is derived from the gene expression measurements using gene set variation analysis (GSVA), gene set enrichment analysis (GSEA), enrichment algorithm, multiscale embedded gene co-expression network analysis (MEGENA), weighted gene co-expression network analysis (WGCNA), differential expression analysis, Z-score, log2 expression analysis, or any combination thereof.

51. The method of any one of claims 47 to 49, wherein the patient data set is derived from the gene expression measurements using GSVA.

52. The method of any one of claims 47 to 51, wherein the machine learning model is trained using linear regression, logistic regression (LOG), Ridge regression, Lasso regression, elastic net (EN) regression, support vector machine (SVM), gradient boosted machine (GBM), k nearest neighbors (kNN), generalized linear model (GLM), naïve Bayes (NB) classifier, neural network, Random Forest (RF), deep learning algorithm, linear discriminant analysis (LDA), decision tree learning (DTREE), adaptive boosting (ADB), Classification and Regression Tree (CART), hierarchical clustering, or any combination thereof.

53. The method of any one of claims 47 to 52, wherein the method classifies whether the patient has cancer: (i) with an accuracy of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (ii) with a sensitivity of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iii) with specificity of at least about 80%, at least about 85%, at least about 90%, at least about 91%,Attorney Docket No.225234-718601 / PCT at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (iv) with a positive predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%; (v) with a negative predictive value of at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, or more than about 99%..

54. The method of any one of claims 47 to 53, wherein the machine learning model has a receiver operating characteristic (ROC) curve with an Area-Under-Curve (AUC) of at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, at least about 0.99, or more than about 0.99, for classifying whether the patient has cancer.

55. The method of any one of claims 47 to 54, wherein the biological sample comprises a blood sample, isolated peripheral blood mononuclear cells (PBMCs), a tissue biopsy sample, or any derivative thereof.

56. The method of any one of claims 47 to 55, wherein the cancer is a solid cancer; optionally wherein the solid cancer is adrenal cancer, anal cancer, bile duct cancer, bladder cancer, bone cancer, brain cancer, breast cancer, carcinoid cancer, cervical cancer, colorectal cancer, esophageal cancer, eye cancer, gallbladder cancer, gastrointestinal stromal tumor, germ cell cancer, head and neck cancer, kidney cancer, liver cancer, lung cancer, nasal cavity and paranasal sinus cancer, nasopharyngeal cancer, neuroblastoma, neuroendocrine cancer, oral cancer, oropharyngeal cancer, ovarian cancer, pancreatic cancer, pediatric cancer, penile cancer, pituitary cancer, prostate cancer, skin cancer, soft tissue cancer, spinal cord cancer, stomach cancer, testicular cancer, thymus cancer, thyroid cancer, ureteral cancer, uterine cancer, vaginal cancer, metastatic renal cell carcinoma, melanoma, carcinoma, a sarcoma, or vulvar cancer.

57. The method of any one of claims 47 to 55, wherein the cancer is a blood cancer; optionally wherein the blood cancer is leukemia, non-Hodgkin lymphoma, Hodgkin lymphoma, an AIDS- related lymphoma, multiple myeloma, plasmacytoma, post-transplantation lymphoproliferative disorder, or Waldenstrom macroglobulinemia.

58. The method of any one of claims 47 to 57, further comprising selecting, recommending and / or administering a treatment to the patient, when the method classifies that the patient has cancer.

59. The method of claim 58, wherein the treatment comprises chemotherapy, radiation, surgery and / or any combination thereof.

Images