Method for identifying benign and malignant thyroid nodules

Abstract

The invention discloses a method for identifying benign and malignant thyroid nodules. The invention provides application of a reagent for detecting the expression level of genes in a sample of a subject to preparation of a kit for diagnosing or predicting thyroid cancer of the subject or identifying benign and malignant thyroid nodules of the subject, wherein the genes comprise one or more genes selected from the group consisting of TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1. The invention also provides a corresponding device and a kit.

Description

[0001] This application claims the priority benefit of the Chinese invention patent application with application date: August 17, 2023, application number: 202311044213.X, and invention name: "Method for identifying benign and malignant thyroid nodules". Technical Field

[0002] The present invention relates to a method or a device for identifying benign or malignant thyroid nodules. Background Art

[0003] In recent years, thyroid nodules have become more common in the population. The proportion of thyroid nodules found by palpation and ultrasound in the population is 7% and 19-67%, respectively, but only 8-16% of thyroid nodules are thyroid cancer. Thyroid cancer is the most common malignant tumor of the endocrine system, ranking ninth in the global cancer incidence. 90% of thyroid cancers are differentiated thyroid cancer (DTC), including subtypes such as papillary carcinoma, follicular carcinoma, and Hürthle cell carcinoma. Another tissue type of thyroid cancer is differentiated thyroid cancer (anaplastic thyroidcarcinoma, ATC). DTC usually has a good prognosis, while the median growth period of ATC patients is only 5 months, with a 1-year survival rate of 20%. Therefore, preoperative diagnosis of thyroid nodules is critical for the selection of treatment options, especially whether surgery is needed and whether to expand the scope of surgery.

[0004] Ultrasound of thyroid nodules, especially ultrasound-guided thyroid fine needle aspiration biopsy (FNA), can effectively differentiate benign from malignant thyroid nodules. However, there are still two types of FNA cytological diagnoses that cannot clearly identify benign or malignant nodules: "atypical hyperplasia", accounting for 2%-18% of FNA; and "follicular hyperplasia or tendency to follicular adenoma", accounting for 2%-25% of FNA.

[0005] The Cancer Genome Atlas (TCGA) divided PTC into BRAFV600E-like and RAS-like according to molecular characteristics in its study of PTC. The mutated genes of RAS-like PTC include NRAS, HRAS, KRAS, EIF1AX, PPARG fusion gene, THADA fusion gene, non-V600E mutated BRAF, etc. RAS-like PTCs are mostly follicular and encapsulated. Many RAS-like PTCs have been reclassified as malignant non-invasive follicular tumors with papillary nuclear features and well-differentiated tumors of uncertain malignant potential. The malignancy rate of thyroid tumors carrying N / H / KRAS mutations reported in the literature is 25% to 90%. The sample size of literature reports on EIF1AX, PPARG fusion gene, THADA fusion gene, and non-V600E mutations of BRAF is limited, but the malignancy percentage is not high.

[0006] When guiding the differentiation of benign and malignant thyroid tumors, RAS-like mutations, whether used as "rule out" or "rule in" markers, lead to extremely high false positive and false negative errors.

[0007] The most common molecular alteration in FNA cytology samples with indeterminate results is RAS mutation. In the past decade, there has been controversy over the use of RAS-like mutations as molecular markers for the differentiation of benign and malignant thyroid nodules: in 2013, a study suggested that thyroid nodules with RAS mutations indicated low-risk thyroid cancer and surgery was recommended

[45] ; in 2019, a study found that the percentage of malignancy in RAS mutation nodules was 25%, suggesting that the detection of RAS mutations in thyroid nodules should be considered an unclear molecular result

[11] ; ThyroSeq tends to consider RAS as a malignant molecular result. The gene classifier (GC) of ThyroSeq V3 can assign weights to the degree of malignancy relevance of the detected mutations and calculate the GC score to determine the benign and malignant nature of the thyroid nodules, but the specific weight of RAS mutations is not mentioned in the literature

[24] ; Afirma GSC uses a machine learning model to differentiate benign and malignant thyroid nodules at the transcriptional level

[27] . The performance characteristics of ThyroSeq and Afirma are high negative predictive value (NPV) and low positive predictive value (PPV), which is similar to the low malignant percentage of RAS-like variants. Therefore, the differentiation of benign and malignant RAS-like thyroid nodules is the focus and difficulty of the molecular diagnosis of thyroid nodules.

[0008] The ThyroSeq series began to detect RAS-like mutations such as N / H / KRAS in V1. In the latest ThyroSeq V3, weights are assigned according to the malignancy correlation of the detected mutations, and the GC score is calculated for the differentiation of benign and malignant thyroid nodules. Although the malignancy correlation weights of mutations are not disclosed, in ThyroSeq V3, tumors with the same mutations, including RAS-like mutations, will obtain equal GC scores so that the benign and malignant differentiation results are consistent, that is, the judgment results are all malignant tumors or all benign and malignant tumors, and false positive or false negative errors cannot be avoided.

[0009] Afirma GEC uses machine learning algorithms to train on RNA microarray results of benign and malignant thyroid tumor samples, and implements a model based on the expression of 167 genes to judge thyroid nodules as "benign" or "suspected malignant". In addition, Afirma has improved the new version of GSC to RNAseq detection, and the number of genes included in the benign and malignant differentiation model has increased to 10,196, with 1,115 core genes. The application objects of GEC / GSC also include RAS-like thyroid tumors, but it does not point out the diagnostic value of this subtype of thyroid tumors, and has the defects of using many genes in the machine learning model and high RNAseq sequencing costs.

[0010] Papillary thyroid carcinoma (PTC) is the most common pathological type of thyroid cancer, accounting for about 85% to 90% of all thyroid cancers [3]. BRAF V600E is the most common driver mutation in PTC, accounting for 30-90%, followed by H / N / KRAS mutation, RET / PTC recombination, etc. [6,9]. Most of these mutations are mutually exclusive. The vast majority of mutations can cause activation of the MAPK pathway. The progression of PTC is associated with the accumulation of mutations in genes such as TERT, TP53, PIK3CA, and AKT1 [10,11]. In the TCGA study on PTC, PTC was divided into BRAF V600E-like and RAS-like based on the transcriptome characteristics of tumor cells [9]. The thyroid differentiation score (TDS) was calculated by the expression levels of 9 genes related to thyroid iodine metabolism in transcriptome sequencing (RNAseq) to indicate the differentiation level of papillary thyroid carcinoma. BRAF V600E-like has a classic papillary morphology, a high level of MAPK signaling pathway activation, and a low differentiation index; RAS-like is mostly follicular growth pattern, encapsulation, a higher proportion of RAS mutations, a low level of MAPK signaling pathway activation, and a higher differentiation index. Currently, many RAS-like PTCs have been reclassified as malignant non-invasive follicular tumors with papillary nuclear features and well-differentiated tumors of uncertain malignant potential.

[0011] Therefore, there is an urgent need for methods that can improve the accuracy of diagnosing benign and malignant thyroid nodules. Summary of the invention

[0012] The present invention provides a method for diagnosing or predicting thyroid cancer in a subject or identifying benign or malignant thyroid nodules in a subject by using the expression levels of one or more genes selected from TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

[0013] In a first aspect, the present invention provides the use of a reagent for detecting the expression level of a gene in a sample of a subject in the preparation of a kit for diagnosing or predicting thyroid cancer in the subject or identifying the benign or malignant nature of the subject's thyroid nodules, wherein the gene comprises one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

[0014] In some embodiments, the expression level of the gene is the level of mRNA transcribed from the gene. In other embodiments, the expression level of the gene is the level of protein translated from the gene.

[0015] In some embodiments, the sample is a tumor tissue, preferably, the tumor tissue is a liquid nitrogen frozen tissue sample, a FNA (fine needle aspiration) sample and / or a FFPE sample. In other embodiments, the sample is blood.

[0016] In some embodiments, the subject is a human.

[0017] In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, and SLC5A5. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, and TG. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, and DIO2. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, and DIO1. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, and GLIS3. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, and SLC26A4.

[0018] In some embodiments, the reagent comprises a primer molecule that hybridizes with the mRNA sequence or its fragment or its transformed variant, and the primer molecule can reverse transcribe and amplify cDNA. Preferably, the reagent also comprises a probe molecule that hybridizes with the cDNA sequence or its fragment or its transformed variant.

[0019] In some embodiments, the diagnosis comprises: obtaining a score by calculation, and diagnosing or predicting the subject's thyroid cancer or identifying the benign or malignant thyroid nodules of the subject according to the score. In some embodiments, the calculation is performed by a random forest model.

[0020] In a second aspect, the present invention provides a kit for diagnosing thyroid cancer in the subject or identifying benign or malignant thyroid nodules in the subject, the kit comprising reagents for detecting the expression levels of genes in the subject's sample, wherein the genes comprise one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

[0021] In some embodiments, the kit includes a chip, and / or primer molecules that hybridize with the mRNA sequence transcribed from the gene or its fragments or their transformed variants, the primer molecules can reverse transcribe and amplify cDNA, and / or probe molecules that hybridize with the cDNA sequence or its fragments or their transformed variants.

[0022] In a third aspect, the present invention provides the use of gene expression levels in constructing a computational model for diagnosing or predicting thyroid cancer in the subject or identifying benign or malignant thyroid nodules in the subject, wherein the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

[0023] In some embodiments, the input variable of the computational model is the expression level of the gene, preferably the level of mRNA transcribed from the gene, more preferably the level of mRNA transcribed from the gene in tumor tissue.

[0024] In some embodiments, the calculation is a model that uses the constructed thyroid cancer diagnosis model to obtain a diagnosis result indicating whether the subject's thyroid nodule is benign or malignant. In some embodiments, the thyroid cancer diagnosis model is a random forest model.

[0025] In a fourth aspect, the present invention provides a device for diagnosing or predicting thyroid cancer or distinguishing benign and malignant thyroid nodules, the device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the program:

[0026] (1) obtaining the expression level of genes in a sample of the subject, wherein the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1,

[0027] (2) inputting the expression level of the gene into a thyroid cancer diagnosis model based on a random forest algorithm, and

[0028] (3) Diagnose thyroid cancer or identify thyroid nodules as benign or malignant based on the output results of the thyroid cancer diagnostic model.

[0029] In some embodiments, step (1) includes detecting the expression level of the gene in the sample, preferably includes detecting the level of mRNA transcribed from the gene in the sample, and more preferably includes detecting the level of mRNA transcribed from the gene in tumor tissue, and the detection includes but is not limited to: RNA extraction, reverse transcription, PCR, and DNA sequencing.

[0030] In a fifth aspect, the present invention provides a device for diagnosing or predicting thyroid cancer or distinguishing between benign and malignant thyroid nodules, comprising:

[0031] A data acquisition unit is configured to acquire sample data including tissue samples of multiple normal persons and sample data including tissue samples of multiple thyroid nodule patients as a first training data set, wherein the first training data set includes a first number of sample data, and each of the sample data includes abundance data of gene expression of a second number of types;

[0032] Diagnostic model construction unit, including:

[0033] a feature screening unit, configured to perform dimensionality reduction processing on the first training data set using a lasso regression algorithm to obtain a second training data set, wherein the second training data includes a first number of dimensionality reduction sample data, and each of the dimensionality reduction sample data includes abundance data of gene expression of a third number of types screened from the second number of types of genes, preferably, important genes related to benign and malignant are screened by variance filtering, mutual information method screening, and model method screening, and preferably, the model screening method is Recursive feature elimination (RFE);

[0034] a model construction unit configured to use a random forest algorithm to train multiple decision trees using the second training data set to construct a random forest model as a thyroid cancer diagnosis model; and

[0035] A diagnostic unit is configured to obtain a diagnostic result indicating whether a thyroid nodule of a subject is benign or malignant using the constructed thyroid cancer diagnostic model.

[0036] In a sixth aspect, the present invention provides a computer storage medium storing instructions, which, when executed by one or more computers, enables the one or more computers to implement the apparatus described in the fourth aspect or the fifth aspect of the present invention.

[0037] Other advantages, features, and utilities of the present invention will be apparent from the detailed description of certain exemplary, non-limiting embodiments; the accompanying drawings; the non-limiting examples; and the claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A schematic diagram of constructing and validating a diagnostic model for benign and malignant thyroid nodules is shown.

[0039] Figure 2 The model contribution of 9 genes in the RAS-like thyroid nodule benign and malignant diagnosis model is shown.

[0040] Figure 3 The figure shows the diagnostic results of 41 RAS-like thyroid nodules using the RAS-like thyroid nodule benign and malignant diagnostic model ("0" means the model diagnoses them as benign, and "1" means the model diagnoses them as malignant), molecular characteristics, pathological types, and the expression of the 9 genes included in the model in each sample. DETAILED DESCRIPTION OF THE INVENTION

[0042] The above features and advantages of the present invention and additional features and advantages will be more clearly understood hereinafter through the following detailed description of the embodiments in conjunction with the accompanying drawings.

[0043] The embodiments described herein with reference to the accompanying drawings are illustrative, exemplary, and are used for a general understanding of the present invention. The embodiments should not be construed as limiting the scope of the present invention. The same or similar elements and elements with the same or similar functions are represented by the same reference numerals throughout the specification.

[0044] Unless otherwise mentioned or defined, all terms used have their ordinary meanings in the art that are clear to those skilled in the art. For example, reference is made to standard manuals, such as Leuenberger, HGW, Nagel, B. and Klbl, H. eds., "A multilingual glossary of biotechnological terms: (IUPAC Recommendations)", Helvetica Chimica Acta (1995), CH-4010 Basel, Switzerland; Sambrook et al., "Molecular Cloning: A Laboratory Manual" (2nd Ed.), Vols. 1-3, Cold Spring Harbor Laboratory Press (1989); F. Ausubel et al., eds., "Current protocols in molecular biology", Green Publishing and Wiley InterScience, New York (1987); Roitt et al., "Immunology (6th Ed.), Mosby / Elsevier, Edinburgh (2001); and Janeway et al., "Immunobiology" (6th Ed.), Garland Science Publishing / Churchill Livingstone, New York (2005), and the general background art cited above.

[0045] definition

[0046] As used herein, the singular forms "a", "an", and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "an antibody" includes a plurality of antibodies and in some embodiments reference to "an antibody" includes a plurality of antibodies, and so forth.

[0047] Unless otherwise stated or defined, the term “comprise” and variations such as “include” and “comprising” will be understood to mean the inclusion of stated elements or steps or groups of elements or steps but not the exclusion of any other elements or steps or groups of elements or steps.

[0048] In the study of papillary thyroid carcinoma (PTC) by The Cancer Genome Atlas (TCGA), PTC was divided into BRAF-like and RAS-like according to molecular characteristics. The mutated genes of RAS-like PTC include NRAS, HRAS, KRAS, EIF1AX, PPARG fusion gene, THADA fusion gene, IGF2BP3 fusion gene, non-V600E mutated BRAF, etc. The present invention extends this concept from PTC to all benign and malignant thyroid nodules, and defines thyroid nodules with detected RAS-like mutant genes as "RAS-like thyroid nodules".

[0049] As used herein, the term "RAS-like thyroid nodules" refers to thyroid nodules including NRAS, HRAS, KRAS, EIF1AX, PPARG fusion gene, THADA fusion gene, IGF2BP3 fusion gene, BRAF with non-V600E mutation, etc. As used herein, the term "non-RAS-like thyroid nodules" refers to thyroid nodules other than RAS-like thyroid nodules.

[0050] In the embodiments of the present invention, the detected mutations involve the following mutant genes: RAS, EIF1AX, BRAF non-V600E variants (BRAF other), PPARG fusions, THADA fusions, IGF2BP3 fusions. Therefore, in the context of the embodiments of the present invention, "RAS-like thyroid nodules" include thyroid nodules involving gene mutations such as RAS, EIF1AX, BRAF non-V600E variants (BRAF other), PPARG fusions, THADA fusions, and IGF2BP3 fusions. In the context of the embodiments of the present invention, the term "non-RAS-like thyroid nodules" refers to thyroid nodules other than thyroid nodules including NRAS, HRAS, KRAS, EIF1AX, PPARG fusion genes, THADA fusion genes, IGF2BP3 fusion genes, non-V600E BRAF mutations, etc.

[0051] As used herein, the term "sequence identity" refers to the degree to which two sequences (amino acids or deoxyribonucleotides or ribonucleotides) have identical residues at identical positions after alignment. For example, "an amino acid sequence is X% identical to SEQ ID NO:Y" means that the amino acid sequence is X% identical to SEQ ID NO:Y and is described as X% of the residues in the amino acid sequence being identical to the residues in the sequence disclosed in SEQ ID NO:Y. Typically, such calculations are performed using computer programs. Exemplary programs for comparing and aligning sequence pairs include ALIGN (Myers and Miller, 1988), FASTA (Pearson and Lipman, 1988; Pearson, 1990), and gapped BLAST (Altschul et al., 1997), BLASTP, BLASTN, or GCG (Devereux et al., 1984).

[0052] As used herein, the term "subject" refers to any mammalian subject for whom diagnosis, treatment, or therapy is desired. "Mammal" for therapeutic purposes refers to any animal classified as a mammal, including humans, livestock, and laboratory, zoo, sports, or pet animals such as dogs, horses, cats, cows, sheep, goats, pigs, mice, rats, rabbits, guinea pigs, monkeys, and the like.

[0053] In the inventor's previous study, RAS-like variant genes were detected in 57 thyroid nodules: RAS, EIF1AX, BRAF non-V600E variant (BRAF other), PPARG fusion, IGF2BP3 fusion, which are defined as "RAS-like thyroid nodules" in the present invention. The malignant percentages of N / H / KRAS (NRAS, HRAS and KRAS), EIF1AX, and IGF2BP3 fusions were 29.63%, 11.11%, and 0.00%, respectively, and the overall malignant percentage of RAS-like thyroid nodules was 23.26%. It can be seen that when RAS-like variants are used as molecular markers in the differentiation of benign and malignant thyroid nodules, both benign markers and malignant markers will lead to extremely high false positive errors and false negative errors.

[0054] The present invention uses a machine learning method to extract features related to benign and malignant thyroid nodules (gene screening) based on the thyroid nodule panel gene expression data of thyroid nodules with clear pathological benign and malignant diagnoses, while focusing on the expression of multiple thyroid differentiation-related genes to establish a benign and malignant thyroid nodule classification model (differential diagnosis model), which is applied to the differentiation of benign and malignant RAS-like mutation nodules.

[0055] In a first aspect, the present invention provides the use of a reagent for detecting the expression level of a gene in a sample of a subject in the preparation of a kit for diagnosing or predicting thyroid cancer in the subject or identifying the benign or malignant nature of the subject's thyroid nodules, wherein the gene comprises one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

[0056] In some embodiments, the expression level of the gene is the level of mRNA transcribed from the gene. In other embodiments, the expression level of the gene is the level of protein translated from the gene.

[0057] In some embodiments, the sample is a tumor tissue, preferably, the tumor tissue is a liquid nitrogen frozen tissue sample, a FNA (fine needle aspiration) sample and / or a FFPE sample. In other embodiments, the sample is blood.

[0058] In some embodiments, the subject is a human.

[0059] In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, and SLC5A5. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, and TG. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, and DIO2. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, and DIO1. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, and GLIS3. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, and SLC26A4.

[0060] In some embodiments, the reagent comprises a primer molecule that hybridizes with the mRNA sequence or its fragment or its transformed variant, and the primer molecule can reverse transcribe and amplify cDNA. Preferably, the reagent also comprises a probe molecule that hybridizes with the cDNA sequence or its fragment or its transformed variant.

[0061] In some embodiments, the diagnosis comprises: obtaining a score by calculation, and diagnosing or predicting the subject's thyroid cancer or identifying the benign or malignant thyroid nodules of the subject according to the score. In some embodiments, the calculation is performed by a random forest model.

[0062] In a second aspect, the present invention provides a kit for diagnosing thyroid cancer in the subject or identifying benign or malignant thyroid nodules in the subject, the kit comprising reagents for detecting the expression levels of genes in the subject's sample, wherein the genes comprise one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

[0063] In some embodiments, the kit includes a chip, and / or primer molecules that hybridize with the mRNA sequence transcribed from the gene or its fragments or their transformed variants, the primer molecules can reverse transcribe and amplify cDNA, and / or probe molecules that hybridize with the cDNA sequence or its fragments or their transformed variants.

[0064] In some embodiments, the expression level of the gene is the level of mRNA transcribed from the gene. In other embodiments, the expression level of the gene is the level of protein translated from the gene.

[0065] In some embodiments, the sample is a tumor tissue, preferably, the tumor tissue is a liquid nitrogen frozen tissue sample, a FNA (fine needle aspiration) sample and / or a FFPE sample. In other embodiments, the sample is blood.

[0066] In some embodiments, the subject is a human.

[0067] In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, and SLC5A5. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, and TG. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, and DIO2. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, and DIO1. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, and GLIS3. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, and SLC26A4.

[0068] In some embodiments, the reagent comprises a primer molecule that hybridizes with the mRNA sequence or its fragment or its transformed variant, and the primer molecule can reverse transcribe and amplify cDNA. Preferably, the reagent also comprises a probe molecule that hybridizes with the cDNA sequence or its fragment or its transformed variant.

[0069] In some embodiments, the diagnosis comprises: obtaining a score by calculation, and diagnosing or predicting the subject's thyroid cancer or identifying the benign or malignant thyroid nodules of the subject according to the score. In some embodiments, the calculation is performed by a random forest model.

[0070] In a third aspect, the present invention provides the use of gene expression levels in constructing a computational model for diagnosing or predicting thyroid cancer in the subject or identifying benign or malignant thyroid nodules in the subject, wherein the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

[0071] In some embodiments, the input variable of the computational model is the expression level of the gene, preferably the level of mRNA transcribed from the gene, more preferably the level of mRNA transcribed from the gene in tumor tissue.

[0072] In some embodiments, the calculation is a model that uses the constructed thyroid cancer diagnosis model to obtain a diagnosis result indicating whether the subject's thyroid nodule is benign or malignant. In some embodiments, the thyroid cancer diagnosis model is a random forest model.

[0073] In some embodiments, the expression level of the gene is the level of mRNA transcribed from the gene. In other embodiments, the expression level of the gene is the level of protein translated from the gene.

[0074] In some embodiments, the sample is a tumor tissue, preferably, the tumor tissue is a liquid nitrogen frozen tissue sample, a FNA (fine needle aspiration) sample and / or a FFPE sample. In other embodiments, the sample is blood.

[0075] In some embodiments, the subject is a human.

[0076] In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, and SLC5A5. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, and TG. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, and DIO2. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, and DIO1. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, and GLIS3. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, and SLC26A4.

[0077] In some embodiments, the reagent comprises a primer molecule that hybridizes with the mRNA sequence or its fragment or its transformed variant, and the primer molecule can reverse transcribe and amplify cDNA. Preferably, the reagent also comprises a probe molecule that hybridizes with the cDNA sequence or its fragment or its transformed variant.

[0078] In some embodiments, the diagnosis comprises: obtaining a score by calculation, and diagnosing or predicting the subject's thyroid cancer or identifying the benign or malignant thyroid nodules of the subject according to the score. In some embodiments, the calculation is performed by a random forest model.

[0079] In a fourth aspect, the present invention provides a device for diagnosing or predicting thyroid cancer or distinguishing benign and malignant thyroid nodules, the device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the program:

[0080] (1) obtaining the expression level of genes in a sample of the subject, wherein the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1,

[0081] (2) inputting the expression level of the gene into a thyroid cancer diagnosis model based on a random forest algorithm, and

[0082] (3) Diagnose thyroid cancer or identify thyroid nodules as benign or malignant based on the output results of the thyroid cancer diagnostic model.

[0083] In some embodiments, step (1) includes detecting the expression level of the gene in the sample, preferably includes detecting the level of mRNA transcribed from the gene in the sample, and more preferably includes detecting the level of mRNA transcribed from the gene in tumor tissue, and the detection includes but is not limited to: RNA extraction, reverse transcription, PCR, and DNA sequencing.

[0084] In some embodiments, the expression level of the gene is the level of mRNA transcribed from the gene. In other embodiments, the expression level of the gene is the level of protein translated from the gene.

[0085] In some embodiments, the sample is a tumor tissue, preferably, the tumor tissue is a liquid nitrogen frozen tissue sample, a FNA (fine needle aspiration) sample and / or a FFPE sample. In other embodiments, the sample is blood.

[0086] In some embodiments, the subject is a human.

[0087] In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, and SLC5A5. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, and TG. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, and DIO2. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, and DIO1. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, and GLIS3. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, and SLC26A4.

[0088] In some embodiments, the reagent comprises a primer molecule that hybridizes with the mRNA sequence or its fragment or its transformed variant, and the primer molecule can reverse transcribe and amplify cDNA. Preferably, the reagent also comprises a probe molecule that hybridizes with the cDNA sequence or its fragment or its transformed variant.

[0089] In some embodiments, the diagnosis comprises: obtaining a score by calculation, and diagnosing or predicting the subject's thyroid cancer or identifying the benign or malignant thyroid nodules of the subject according to the score. In some embodiments, the calculation is performed by a random forest model.

[0090] In a fifth aspect, the present invention provides a device for diagnosing or predicting thyroid cancer or distinguishing between benign and malignant thyroid nodules, comprising:

[0091] A data acquisition unit is configured to acquire sample data including tissue samples of multiple normal persons and sample data including tissue samples of multiple thyroid nodule patients as a first training data set, wherein the first training data set includes a first number of sample data, and each of the sample data includes abundance data of gene expression of a second number of types;

[0092] Diagnostic model construction unit, including:

[0093] a feature screening unit, configured to perform dimensionality reduction processing on the first training data set using a lasso regression algorithm to obtain a second training data set, wherein the second training data includes a first number of dimensionality reduction sample data, and each of the dimensionality reduction sample data includes abundance data of gene expression of a third number of types screened from the second number of types of genes, preferably, important genes related to benign and malignant are screened by variance filtering, mutual information method screening, and model method screening, and preferably, the model screening method is Recursive feature elimination (RFE);

[0094] a model construction unit configured to use a random forest algorithm to train multiple decision trees using the second training data set to construct a random forest model as a thyroid cancer diagnosis model; and

[0095] A diagnostic unit is configured to obtain a diagnostic result indicating whether a thyroid nodule of a subject is benign or malignant using the constructed thyroid cancer diagnostic model.

[0096] In some embodiments, the expression level of the gene is the level of mRNA transcribed from the gene. In other embodiments, the expression level of the gene is the level of protein translated from the gene.

[0097] In some embodiments, the sample is a tumor tissue, preferably, the tumor tissue is a liquid nitrogen frozen tissue sample, a FNA (fine needle aspiration) sample and / or a FFPE sample. In other embodiments, the sample is blood.

[0098] In some embodiments, the subject is a human.

[0099] In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, and SLC5A5. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, and TG. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, and DIO2. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, and DIO1. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, and GLIS3. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, and SLC26A4.

[0100] In some embodiments, the reagent comprises a primer molecule that hybridizes with the mRNA sequence or its fragment or its transformed variant, and the primer molecule can reverse transcribe and amplify cDNA. Preferably, the reagent also comprises a probe molecule that hybridizes with the cDNA sequence or its fragment or its transformed variant.

[0101] In some embodiments, the diagnosis comprises: obtaining a score by calculation, and diagnosing or predicting the subject's thyroid cancer or identifying the benign or malignant thyroid nodules of the subject according to the score. In some embodiments, the calculation is performed by a random forest model.

[0102] In a sixth aspect, the present invention provides a computer storage medium storing instructions, which, when executed by one or more computers, enables the one or more computers to implement the apparatus described in the fourth aspect or the fifth aspect of the present invention.

[0103] In addition, the present invention also provides a method for diagnosing or predicting thyroid cancer in the subject or identifying benign or malignant thyroid nodules in the subject, the method comprising detecting the expression level of genes in the subject's sample, wherein the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

[0104] In some embodiments, the expression level of the gene is the level of mRNA transcribed from the gene. In other embodiments, the expression level of the gene is the level of protein translated from the gene.

[0105] In some embodiments, the sample is a tumor tissue, preferably, the tumor tissue is a liquid nitrogen frozen tissue sample, a FNA (fine needle aspiration) sample and / or a FFPE sample. In other embodiments, the sample is blood.

[0106] In some embodiments, the subject is a human.

[0107] In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, and SLC5A5. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, and TG. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, and DIO2. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, and DIO1. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, and GLIS3. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, and GLIS3. In some embodiments, the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, and SLC26A4.

[0108] In some embodiments, the reagent comprises a primer molecule that hybridizes with the mRNA sequence or its fragment or its transformed variant, and the primer molecule can reverse transcribe and amplify cDNA. Preferably, the reagent also comprises a probe molecule that hybridizes with the cDNA sequence or its fragment or its transformed variant.

[0109] In some embodiments, the diagnosis comprises: obtaining a score by calculation, and diagnosing or predicting the subject's thyroid cancer or identifying the benign or malignant thyroid nodules of the subject according to the score. In some embodiments, the calculation is performed by a random forest model. Example

[0110] The present invention uses the expression levels of genes in the patient's thyroid nodules and a random forest algorithm to train a diagnostic model, and verifies the accuracy of the diagnostic model on an independent test set.

[0111] Unless explicitly stated otherwise, the experimental protocols or procedures in the examples of the present invention were performed using conventional methods.

[0112] Materials and Methods

[0113] 1. Instruments and Equipment

[0114] (1) High-throughput sequencer Nextseq CN500 (Illumina)

[0115] (2) Qubit TM 4.0 (Invitrogen TM )

[0116] (3) LabChip (PerkinElmer)

[0117] (4) Centrifuge (Eppendorf)

[0118] (5)NanoDrop TM 2000

[0119] (6) Fragmentation instrument (Covaris M220)

[0120] (7) Mini centrifuge (TIANGEN)

[0121] (8) Low temperature homogenizer (TIANGEN)

[0122] (9) Vortex analyzer (VORTEX-GENIE2)

[0123] (10)PCR instrument (Bio-rad S1000 and ABI Veriti)

[0124] (11) Pipette (Eppendorf Research Plus, etc.)

[0125] (12) Dry constant temperature metal bath (Eppendorf, etc.)

[0126] (13) Magnetic rack (Life and TIANGEN)

[0127] (14) Server (Inspur, etc.)

[0128] 2. Reagents and consumables

[0129] (1) xGen TMCustom Probe Set (A-histidine cDNA) (Integrated DNA Technologies)

[0130] (2) xGen TM Custom Probe Set (A-gDNA) (Integrated DNA Technologies)

[0131] (3)xGen TM Hybridization and Wash Kit(Intergrated DNA Technologies)

[0132] (4) xGen TM Universal Blockers-TS Mix(Intergrated DNA Technologies)

[0133] (5)KAPA HiFi HotStart ReadyMix (KAPA)

[0134] (6)KAPA Pure Beads (KAPA)

[0135] (7)KAPA Hyper Prep Kit (KAPA)

[0136] (8)KAPA mRNA Capture Kit (KAPA)

[0137] (9)KAPA RNA HyperPrep Kit (KAPA)

[0138] (10)NextSeq 500 / 550Mid Output Kit(Illumina)

[0139] (11) Agencourt AMPure XP Beads (Beckman)

[0140] (12)RNase,DNase,RNA and DNA Away(Beyotime)

[0141] (13) Nuclease-free water (non-DEPC treated) (Invitrogen TM )

[0142] (14) UltraPure TM DNase / RNase-Free Distilled Water(Invitrogen TM )

[0143] (15)Qubit TM dsDNA HS Detection Kit (Invitrogen TM )

[0144] (16)Qubit TM dsDNA BR detection kit (Invitrogen TM )

[0145] (17)Qubit TM RNA HS Detection Kit (Invitrogen TM )

[0146] (18)Qubit TM RNA IQ Assay Kit (Invitrogen TM )

[0147] (19)24DNA 5K / RNA LABCHIP (PerkinElmer)

[0148] (20)RNA Reagent Kit (PerkinElmer)

[0149] (21)24DNA Extended Range Labchip(PerkinElmer)

[0150] (22)DNA HiSena Rragent Kit (PerkinElmer)

[0151] (twenty three) Unique Dual Index Barcodes

[0152] (24)AllPrep DNA / RNA Micro Kit(QIAGEN)

[0153] (25)GeneRead DNA FFPE Kit(QIAGEN)

[0154] (26)QIAGEN QIAamp DNA FFPE Tissue Kit(QIAGEN)

[0155] (27)QIAGEN QIAamp DNA Mini Kit(QIAGEN)

[0156] (28)RNeasy FFPE Kit(50)(QIAGEN)

[0157] (29) Proteinase K (QIAGEN)

[0158] (30) RNase A (QIAGEN)

[0159] (31)Qubit TM Assay Tubes (Invitrogen TM )

[0160] (32) 10 / 100 / 200ul low-retention pipette tip (AXYGEN)

[0161] (33)DNA loBind Tube 1.5ml(eppendorf)

[0162] (34) 0.2 / 0.5 / 1.5ml Tube low adsorption (AXYGEN)

[0163] (35)0.2ml PCR Tube Strips(eppendorf)

[0164] (36)microTUBE Snap-Cap Case(covaris)

[0165] (37) 0.01 / 0.1 / 0.2ml low adsorption filter tip (AXYGEN)

[0166] (38) Full skirt 96-well plate (AXYGEN)

[0167] (39)RC-L250 / 10 200ul LTSTM TIPS(RAININ)

[0168] (40)KW32- Wipe paper

[0169] (41) 1 / 2 inch cryo box (Wangge)

[0170] (42) 1M NaOH (Fluka)

[0171] (43) 20(Sigma)

[0172] (44) Sodium hypochlorite solution (Sigma)

[0173] (45) DL-dithiothreitol solution (Sigma)

[0174] 3. Design and synthesis of panel probes for thyroid gland

[0175] The A-Jie Panel achieves targeted enrichment through hybridization capture of biotinylated probes with target gene regions for subsequent high-throughput sequencing. The probe is essentially a 5'-biotin-labeled 20-120nt oligonucleotide that is complementary to the target gene region. The A-Jie Panel contains two sets of probes, gDNA probes for targeted capture of genomic DNA (gDNA) libraries and cDNA probes for targeted capture of messenger RNA (mRNA) reverse transcriptase (cDNA) libraries. The gDNA probe is used to detect substitutions, insertions and deletions, and copy number variations (CNVs) of the target gene, while the cDNA probe is used to detect fusion variations and expression levels of the target gene. The probe distribution principles are as follows:

[0176] (1) The principles of gDNA probe distribution are as follows: ① Probes are distributed at the junction of exons and introns to be specific for gDNA capture; ② Probes are distributed at the gene SNP sites where CNVs need to be detected to detect copy number variations (CNVs). ③ The (CDS) region of the target gene should achieve 2x uniform coverage. After completing the above two probe designs, uniform probes are supplemented for areas not covered by 2x probes.

[0177] (2) The cDNA probe distribution principles are as follows: ① Distribute probes within the range of 20-120 nt at the edge of the exon to enhance the ability to capture new opponent genes; ② Distribute probes across exons to be specific for cDNA capture and avoid gDNA interference; ③ The CDS region of the target gene should achieve 2x uniform coverage. After completing the above two probe designs, uniform probes should be supplemented to the areas not covered by 2x probes.

[0178] The probe for the A-junction Panel uses the xGen TM Custom hybridization capture probe set technology services.

[0179] 4. Sample nucleic acid extraction

[0180] 1.2.4.1 DNA / RNA extraction from liquid nitrogen frozen tissues and FNA samples

[0181] DNA and RNA were extracted from liquid nitrogen frozen tissues and FNA samples using the AllPrep DNA / RNA micro kit (QIAGEN).

[0182] (1) Liquid nitrogen cryopreserved tissue sample collection: Prepare 2.5 ml screw cap tubes, add 550 ul RLT plus lysis buffer and 100 ul 2 mm grinding beads (TIANGEN) to each tube. Take out the liquid nitrogen cryopreservation tube containing the frozen tissue, pour out the tissue on a clean film glove, cut off a mustard-sized sample as quickly as possible, and put it into RLT plus lysis buffer (the amount can be increased according to the situation of the tissue sample). Immediately put the remaining tissue back into the liquid nitrogen tube and throw it into liquid nitrogen. Continue permanent cryopreservation.

[0183] (2) FNA tissue sample collection: Prepare 2.5 ml screw-cap tubes and add 550 ul of RLT plus lysis solution to each tube for FNA puncture. After the fine needle aspiration is completed under B-ultrasound guidance, the puncture needle is irrigated in RLT plus lysis solution for 3-5 times. FNA exfoliated cell precipitation can be seen in the lysis solution.

[0184] (3) Use Tiangen cryogenic homogenizer at -20°C, 45 s, 4 cycles to prepare liquid nitrogen frozen tissue sample / FNA sample homogenate for subsequent DNA and RNA extraction.

[0185] (4) Prepare 250ul of 70% ethanol in a 1.5ml centrifuge tube, take 250ul of the homogenate and mix with 250ul of 70% ethanol by pipetting for RNA extraction.

[0186] (5) The remaining homogenate was used to extract DNA. 250 μl of the homogenate was added to the DNA column and centrifuged at 10,000 rpm for 15 seconds.

[0187] (6) Add 500ul of the mixed liquid in step (4) to the RNA column, centrifuge at 10,000rpm for 15s, and discard the liquid. Add 700u of RW1 buffer, centrifuge at 10,000rpm for 15s, and discard the liquid. Add 500ul of RPE buffer, centrifuge at 10,000rpm for 15s, and discard the liquid. Add 500ul of 80% ethanol, centrifuge at 10,000rpm for 120s. Discard the collection tube and replace it with a new one.

[0188] (7) Open the lid and centrifuge at 10,000 rpm for 6 min to dry the column membrane. Discard the collection tube and replace it with a 1.5 ml RNase-free tube.

[0189] (8) Add 30u RNase-free water, incubate at room temperature for 10min, and centrifuge at maximum speed for 1min to collect the RNA eluate. The column membrane in the kit has a small diameter and a rubber ring around it. When eluting nucleic acids, make sure that the eluate drops in the center of the column membrane and spreads evenly.

[0190] (9) Add 500ul of AW1 buffer to the DNA column that has been centrifuged in step (5), centrifuge at 10,000rpm for 15s, and discard the liquid. Add 500ul of AW2 buffer, centrifuge at 10,000rpm for 15s, discard the collection tube, and replace it with a new collection tube.

[0191] (10) Maximum speed for 2 min to ensure that the column membrane is dry. Change to a 1.5 ml EP tube.

[0192] (11) Preheat EB elution buffer at 70°C, add 30ul EB, incubate at room temperature for 10min, and elute at maximum speed for 1min. Then add another 30ul EB and centrifuge at maximum speed for 1min to collect the DNA eluate, totaling 60ul.

[0193] (12) Use Nanodrop to preliminarily test the DNA and RNA concentrations and A260 / 280. A260 / 280 between 1.8 and 2.0 is considered acceptable.

[0194] 1.2.4.2 DNA extraction from FFPE samples

[0195] Using GeneRead TM DNA was extracted from FFPE samples using DNA FFPE kit (QIAGEN).

[0196] (1) Use a surgical blade to scrape the paraffin tissue of the tumor area on the slide into a centrifuge tube and centrifuge at maximum speed for 5 minutes so that the paraffin tissue accumulates at the bottom of the tube.

[0197] (2) Add 320 μl of Deparaffinization solution to each tube, vortex to mix for 10 seconds, and centrifuge in a handheld centrifuge.

[0198] (3) Heat the mixture in a water bath at 56°C for 2 h.

[0199] (4) After the water bath, add 55ul RNase-free water, 25ul FTB buffer, and 20ul proteinase K in sequence.

[0200] (5) Digest in a water bath at 56°C for 12 h.

[0201] (6) Incubate in a metal bath or water bath at 90°C for 1 h.

[0202] (7) After instant centrifugation, remove the lower layer of transparent liquid and transfer it to another 1.5 ml centrifuge tube.

[0203] (8) Add 115ul RNase-free water and 35ul UNG and vortex to mix.

[0204] (9) Incubate in a water bath at 50°C for 1 h. Centrifuge briefly, add 2 μl of RNase-A, vortex to mix, centrifuge briefly, and let stand at room temperature for 2 min.

[0205] (10) Add 250 μl of AL buffer, vortex mix, centrifuge briefly, then add 250 μl of anhydrous ethanol, vortex mix, and centrifuge briefly.

[0206] (11) Transfer 700 μl of liquid to the column (Qiamp MinElute Column), centrifuge at maximum speed for 1 min, discard the liquid in the collection tube, and repeat the centrifugation until all the liquid is transferred.

[0207] (12) Add 500ul AW1 buffer, centrifuge at maximum speed for 1 min, and discard the liquid in the collection tube. Add 500ul AW2 buffer, centrifuge at maximum speed for 1 min, and discard the liquid in the collection tube. Add 250ul anhydrous ethanol, centrifuge at maximum speed for 1 min, and change the collection tube.

[0208] (13) Centrifuge at maximum speed for 1 min to dry the column membrane. Discard the collection tube and place the column in another 1.5 ml centrifuge tube.

[0209] (14) Add 20-40 μl ATE to the membrane in the center of the column, incubate at room temperature for 10 min, and centrifuge at maximum speed for 1 min.

[0210] (15) Use Nanodrop to preliminarily test DNA concentration and A260 / 280. A260 / 280 between 1.8 and 2.0 is considered acceptable.

[0211] 1.2.4.3 RNA extraction from FFPE samples

[0212] Using QIAGEN FFPE kit was used to extract RNA from FFPE samples.

[0213] (1) Preparation: Use a sterile disposable RNase-free syringe to inject 550ul of RNase-free water into the vial and mix thoroughly by inverting. Add anhydrous ethanol to buffer RPE and shake well.

[0214] (2) Use a scalpel to scrape the tumor tissue off the slide and place it into a 1.5 ml centrifuge tube and centrifuge.

[0215] (3) Add 160 μl of dewaxing solution and centrifuge briefly to make the flocculent sample sink to the bottom of the tube as much as possible;

[0216] (4) Incubate at 56°C for 3 min and cool to room temperature (15-25°C). If too little dewaxing solution is added or the sample contains too much paraffin, the dewaxing solution may become waxy or solid after cooling. In this case, add 160 μl of dewaxing solution, incubate at 56°C for another 3 min, and then cool to room temperature.

[0217] (5) Add 150 μl of buffer PKD and mix by vortexing;

[0218] (6) Centrifugation: 11000 rpm (10000 rcf) for 1 min;

[0219] (7) Add 10 μl of proteinase K to the lower clear liquid phase and mix gently with a pipette tip;

[0220] (8) Incubate at 56°C for 15 min, then at 80°C for 15 min. Try to use a heating block with automatic mixing function. If there is only one heating block, cool the sample to room temperature after incubation at 56°C, and then incubate again when the heating block is raised to 80°C.

[0221] (9) Transfer the lower colorless liquid phase to a new 2 ml centrifuge tube;

[0222] (10) Incubate on ice for 3 min, centrifuge at 13500 rpm (20000 rcf) for 15 min;

[0223] (11) Transfer the supernatant to a new centrifuge tube. Be careful not to suck up the precipitate. You can change to a 10ul pipette tip to suck out as much supernatant as possible.

[0224] (12) Add 16 μl DNase Booster Buffer and 10 μl DNase I working solution, invert to mix, and centrifuge briefly;

[0225] (13) Incubate at room temperature for 15 min;

[0226] (14) Add 320 μl of buffer RBC and mix the solution by pipetting;

[0227] (15) Add 720 μl of anhydrous ethanol (100%) and mix by pipetting. Do not centrifuge and proceed to step 17 immediately. There may be precipitation after adding ethanol, but it will not affect the extraction results;

[0228] (16) Transfer 700ul of sample including the precipitate to a centrifuge column (provided) placed in a 2ml collection tube. Gently cover the column and centrifuge at ≥10,000rpm (≥8,000rcf) for 15s. Discard the filtrate and keep the collection tube.

[0229] (17) Repeat the previous step until all samples have passed through the centrifuge column and retain the collection tube;

[0230] (18) Add 500ul of buffer RPE to the centrifuge column, cover the lid lightly, and centrifuge at ≥10000rpm (≥8000rcf) for 15s. Discard the filtrate and keep the collection tube.

[0231] (19) Note: The buffer RPE provided is a concentrated solution. Make sure to add anhydrous ethanol before use.

[0232] (20) Add 500ul of buffer RPE to the centrifuge column again, cover the column cap, and centrifuge at ≥10000rpm (≥8000rcf) for 2min. Discard the filtrate and collection tube.

[0233] (21) Note: After centrifugation, carefully remove the spin column from the collection tube, ensuring that the spin column does not come into contact with the filtrate, otherwise ethanol residue will be produced.

[0234] (22) Transfer the centrifuge column to a new 2 ml collection tube, open the centrifuge column cap, and centrifuge at maximum speed for 5 min. Discard the filtrate and collection tube;

[0235] (23) To avoid damage to the lid and contamination, when placing the centrifuge, there should be at least two empty spaces between adjacent centrifuge columns (each centrifuge can hold a maximum of 8 tubes of samples), and the lid should be placed in the opposite direction of the rotor's rotation.

[0236] (24) Label the 1.5 ml collection tube provided in the kit and write clearly: RNA-FFPE sample number-T or N-patient name.

[0237] (25) Transfer the centrifuge column to a new 1.5 ml collection tube, add 30 μl of RNase-free water directly onto the center of the membrane, without touching the membrane or the yellow rubber ring, cover the lid, incubate at room temperature for 5 min, and centrifuge at maximum speed for 1 min. (If you are concerned about the low total yield, you can add the liquid in the 1.5 ml collection tube after centrifugation to the center of the membrane again, incubate at room temperature for 5 min, and centrifuge again at maximum speed for 1 min.)

[0238] (26) 260 / 280 measured using Nano drop 2000.

[0239] (27) Concentration determination using qubit.

[0240] (28) If the concentration measured by qubit is higher than 250 ng / ul, the rin value can be directly detected by qubit. If the concentration is lower than 250 ng / ul, use 8-tube strips to divide each tube into 3 ul (pay attention to the label information) and store it at -80 together with the RNA sample, waiting for the labchip to detect the rin value.

[0241] 5Library preparation (library construction) and sequencing

[0242] 5.1gDNA library preparation

[0243] (1) DNA was quantified using Qubit 4.0.

[0244] (2) The starting amount of DNA extracted from liquid nitrogen frozen tissue / FNA samples for library construction is 150 ng, and the starting amount of DNA extracted from FFPE samples for library construction is 500 ng.

[0245] (3) Prepare the samples to be used for library construction in a 200-ul centrifuge tube. Take 150 ng or 500 ng of DNA and add nuclease-free water to make up the total volume to 55 ul.

[0246] (4) Prepare Covaris M220 broken tubes. The sticks inside the tubes must not be bent and there must be one stick per tube.

[0247] (5) Turn on the Covaris M220 and open the program. Fill the ultrasonic tank with ultrapure water using the bottle provided with the instrument and close the door. The program will then display whether the filling, door closing, etc. are qualified. If qualified, continue with the subsequent operation.

[0248] (6) Preparation of broken tube samples: prepare the tube opening plate and tube opener that come with the breaking instrument. Place the broken tube into the hole on the tube opening plate, and use a pipette to transfer all 55ul of the sample to be broken into the corresponding marked broken tube: use a 200ul or 100ul range pipette to adjust 60ul, and after absorbing all the samples, insert the pipette vertically through the rubber tube cap from the top of the broken tube, and then draw out all the samples. When drawing out, keep the pipette empty to prevent the sample from being sucked back.

[0249] (7) DNA fragmentation: Fragment the tube and place it into the fragmentation instrument. Select 75w / 195s and start the program. The target range of DNA fragmentation is 150bp-300bp.

[0250] (8) After the procedure is completed, remove the broken tube, wipe it dry with paper towels, centrifuge it briefly, and place it in the hole on the tube opening plate. Use a tube opener to open the broken tube and transfer all the liquid to a 200ul centrifuge tube. The fragmented DNA sample can be stored at 4°C overnight or at -20°C for a long time.

[0251] (9) Use the KAPA kit to complete the gDNA hybridization capture pre-library construction. Take out the End Repair & A-Tailing Buffer and thaw it at room temperature, mix it evenly, and place it on ice for later use. Take out the End Repair & A-Tailing Enzyme and place it on ice to thaw naturally, mix it evenly, and centrifuge it for later use.

[0252] (10) Prepare the End Repair & A-Tailing reaction system in a 200 μl PCR tube placed on ice and complete the reaction on a PCR instrument. The system and procedure are as follows:

[0253] Table 1 End Repair & A-Tailing reaction system

[0254]

[0255] Table 2 End Repair & A-Tailing reaction procedure

[0256]

[0257] (11) Take out The magnetic beads were balanced at room temperature, the Ligation Buffer was taken out and thawed at room temperature, mixed evenly, and placed on ice for later use, the DNA Ligase and adapter were taken out and placed on ice for natural thawing, mixed evenly, and centrifuged instantly for later use. After End Repair & A-Tailing was completed, the adapter ligation reaction was immediately performed.

[0258] (12) Prepare the following linker ligation reaction system:

[0259] Table 3 Linker ligation reaction system

[0260]

[0261] (13) End Repair & A-Tailing After the reaction is completed, take out the PCR reaction tubes and place them on ice. Add 5 μL of connectors to each tube. Unique Dual Index Barcodes.

[0262] Then add 45 μL of the adapter ligation reaction system configured in the previous step, vortex to mix, and centrifuge instantly.

[0263] The PCR tubes were incubated at 20°C for 15 minutes to obtain the adapter ligation products.

[0264] (14) Use Magnetic beads were used to purify the connector-connected product. 88 μL of magnetic beads were added to each tube for 0.8x magnetic bead purification. The purification system was as follows:

[0265] Table 4 Connector ligation product purification system

[0266]

[0267] (15) Mix the purified system thoroughly by pipetting up and down with the pipette tip. Incubate at room temperature for 5-15 minutes to allow the DNA to bind to the magnetic beads. Place the PCR tube on a magnetic rack and incubate until the liquid is clear. Carefully aspirate and discard the supernatant.

[0268] (16) Keeping the PCR tube on the magnetic rack, add 200 μL of 80% ethanol. Incubate the PCR tube on the magnetic rack at room temperature for ≥ 30 seconds and then carefully aspirate and discard the ethanol.

[0269] (17) Repeat the previous step once.

[0270] (18) Take out the PCR tube, centrifuge briefly, and place it on a magnetic rack. Use a 10 μL pipette tip to try to remove all the remaining ethanol, while trying not to disturb the magnetic beads. Dry the magnetic beads at room temperature for 3-5 minutes, or until all the ethanol has evaporated, to prevent over-drying of the magnetic beads, which may lead to a decrease in library yield.

[0271] (19) Remove the PCR tube from the magnetic stand, completely resuspend the magnetic beads in 25 μL of nuclease-free water, and incubate at room temperature for 2 min to elute the DNA from the magnetic beads.

[0272] (20) Place the PCR tube on a magnetic rack and incubate until the liquid becomes clear. Transfer 20 μL of the clear supernatant to a new PCR tube to obtain the adapter-ligated library for the next library amplification reaction.

[0273] (21) Take out 2X KAPA HiFi HotStart ReadyMix and 10X Library Amplification Primer Mix, place them on ice to thaw, mix them evenly, and centrifuge them for later use. Prepare the library amplification reaction system. The library amplification reaction system and reaction procedure are as follows:

[0274] Table 5 Library amplification reaction system

[0275]

[0276] Table 6 Library amplification reaction program

[0277]

[0278] (22) The amplified library can be stored at 4°C or -20°C and purified within 72 hours.

[0279] (23) Configure the purification system according to the following table and perform 1X magnetic bead purification:

[0280] Table 7 Amplification library purification system

[0281]

[0282]

[0283] (24) Mix the purified system thoroughly by pipetting up and down with the pipette tip and incubate at room temperature for 5-15 minutes to allow the DNA to bind to the magnetic beads. Place the PCR tube on a magnetic rack and incubate until the liquid is clear. Carefully aspirate and discard the supernatant.

[0284] (25) Keeping the PCR tube on the magnetic rack, add 200 μL of 80% ethanol. Incubate the PCR tube on the magnetic rack at room temperature for ≥ 30 seconds and then carefully aspirate and discard the ethanol.

[0285] (26) Repeat the previous step once.

[0286] (27) Take out the PCR tube, centrifuge briefly, and place it on a magnetic rack. Use a 10 μL pipette tip to try to remove all the remaining ethanol, while minimizing the disturbance of the magnetic beads. Dry the magnetic beads at room temperature for 3-5 minutes, or until all the ethanol has evaporated, to prevent over-drying of the magnetic beads, which may result in a decrease in library yield.

[0287] (28) Remove the PCR tube from the magnetic stand, completely resuspend the magnetic beads in 35 μL of nuclease-free water, and incubate at room temperature for 2 min to elute the DNA from the magnetic beads.

[0288] (29) Place the PCR tube on a magnetic stand and incubate until the liquid is clear. Transfer 33 μL of the supernatant to a new 1.5 mL low-adsorption centrifuge tube for storage. Label the centrifuge tube with the library type, number, adapter, and library construction date. Use Qubit 4.0 for quantification. This is the gDNA library, which will be used in the subsequent Panel gDNA probe hybridization capture experiment.

[0289] 1.2.3.2 cDNA library preparation

[0290] cDNA library preparation was performed using the KAPA mRNA Capture Kit and the KAPA RNA HyperPrep Kit.

[0291] (1) RNA was quantified using Qubit 4.0.

[0292] (2) The starting amount of RNA extracted from liquid nitrogen frozen tissue / FNA samples for library construction is 250 ng, and the starting amount of RNA extracted from FFPE samples for library construction is 500 ng.

[0293] (3) Half an hour before library construction, take out the mRNA Capture Beads, mRNA Bead Binding Buffer, mRNA Bead Wash Buffer and Fragment, and Prime and Elute Buffer (2X) and equilibrate them at room temperature.

[0294] (4) Prepare the sample for library construction in a PCR tube. Take 250 ng or 500 ng of RNA and add RNase-free water to make up the total volume to 50 ul.

[0295] (5) Thoroughly resuspend the mRNA Capture Beads and dispense the mRNA capture reagents mRNA Capture Beads, mRNA Bead Binding Buffer, and mRNA Bead Wash Buffer according to the table below, with a 5% loss calculated.

[0296] Table 8 mRNA capture reagent dosage

[0297]

[0298] (6) Configure the mRNA fragmentation system according to the following table:

[0299] Table 9 mRNA fragmentation system

[0300]

[0301] (7) Place the aliquoted mRNA Capture Beads centrifuge tube on a magnetic rack and incubate at room temperature until the solution is clear. Carefully aspirate and discard the supernatant. Add an equal volume of mRNA Bead Binding Buffer. Remove the centrifuge tube from the magnetic rack and resuspend the magnetic beads by pipetting up and down with a low-absorption pipette tip to avoid generating foam.

[0302] (8) Place the centrifuge tube on a magnetic rack and incubate at room temperature until the solution is clear. Carefully aspirate the supernatant and add an equal volume of mRNA Bead Binding Buffer. Remove the centrifuge tube from the magnetic rack and resuspend by pipetting up and down to avoid foaming. 。

[0303] (9) For the RNA sample to be captured, add 50 μL of resuspended mRNA capture beads using a low-adsorption pipette tip and mix thoroughly by gently pipetting up and down.

[0304] (10) Place the PCR tube into the PCR instrument and perform the first mRNA capture. The procedure is as follows:

[0305] Table 10 First mRNA capture procedure

[0306]

[0307] (11) Place the PCR tube containing the mRNA capture magnetic beads and RNA mixture on a magnetic rack and incubate at room temperature until the solution is clear, then discard the supernatant.

[0308] (12) Remove the PCR tube from the magnetic rack and add 200 μL of mRNA Bead Wash Buffer. Use a low-retention pipette tip to mix thoroughly and resuspend.

[0309] (13) Place the PCR tube on a magnetic rack and incubate at room temperature until the solution is clear. Discard the supernatant.

[0310] (14) Add 50 μL of nuclease-free water and resuspend thoroughly by pipetting up and down with a low-absorption pipette tip.

[0311] (15) Place the PCR tube into the PCR instrument and perform the second mRNA capture. The procedure is as follows:

[0312] Table 11 Second mRNA capture

[0313]

[0314] (16) After the second step of mRNA capture, add 50 μL of Bead Binding Buffer and use a low-binding pipette tip to mix thoroughly and resuspend. Incubate the PCR tube at 20°C for 5 minutes.

[0315] (17) Place the PCR tube on a magnetic rack and incubate at room temperature until the solution is clear. Discard the supernatant.

[0316] (18) Remove the PCR tube from the magnetic rack and add 200 μL of mRNA Bead Wash Buffer. Use a low-binding pipette tip to mix thoroughly and resuspend. Place the PCR tube on the magnetic rack and incubate at room temperature until the solution is clear. Aspirate and discard the supernatant.

[0317] (19) Prepare the mRNA elution and fragmentation (Fragment & Elute) system at room temperature as follows:

[0318] Table 12 mRNA Fragment & Elute System

[0319]

[0320] (20) Add 22 μL of the above mixture to thoroughly resuspend the magnetic beads that have captured mRNA. Use a low-absorption pipette tip to mix thoroughly by pipetting up and down. Store in a 4°C refrigerator and proceed to the next step within 24 hours.

[0321] (21) Take out the 1st Strand Synthesis Buffer and KAPA Script and melt them in a 4°C refrigerator. Prepare the 1st Strand Synthesis reaction solution according to the following table:

[0322] Table 13 1st Strand Synthesis System

[0323]

[0324] (22) Take out the mRNA capture magnetic bead resuspension solution temporarily stored at 4°C, and use a low-adsorption pipette tip to pipette up and down to mix thoroughly and resuspend the mRNA capture beads. Then immediately place the PCR tube in the PCR instrument and perform mRNA fragmentation at 85°C / 4min. After completion, immediately place the PCR tube on the magnetic rack and incubate until the liquid is clear. To prevent poly(A)-rich RNA from hybridizing with the capture magnetic beads, the sample cannot be cooled before being placed on the magnetic rack.

[0325] (23) Place the PCR tube on ice and carefully transfer 20 μL of supernatant.

[0326] (24) Add 10 μl of 1st Strand Synthesis reaction solution to each tube, mix by gently pipetting, and then perform first strand synthesis to obtain first strand cDNA. The procedure is as follows:

[0327] Table 14 First-strand synthesis program

[0328]

[0329] (25) The program runs for 40 min. At this time, take out the 2nd Strand Making Buffer and melt it in a 4°C refrigerator.

[0330] (26) 5 min before the end of the first strand synthesis process, prepare the second strand synthesis and A-tailing reaction solution on an ice box. The system is as follows:

[0331] Table 15 Second chain synthesis reaction solution

[0332]

[0333] (27) After the first chain synthesis process is completed, keep the PCR tubes on ice, add 30ul of the second chain synthesis reaction solution to each tube, and mix by gently pipetting up and down. Place the PCR tubes in the PCR instrument and perform the second chain synthesis process as follows:

[0334] Table 16 Second chain synthesis program

[0335]

[0336]

[0337] (28) The program running time of the previous step is about 40 minutes. At this time, take out the Ligation Buffer and thaw it in a 4℃ refrigerator. 5 minutes before the end of the program, prepare the adapter ligation reagent and dilute the adapter. Take it out when preparing the adapter and DNA Ligase reagent. Take out the KAPA Pure Beads and equilibrate them at room temperature for purification after the first ligation. 70ul / library, estimated amount. Take out the PEG / NaCl Solution and equilibrate it at room temperature for purification after the second ligation.

[0338] (29) Configure the mRNA adapter ligation system as shown in the following table:

[0339] Table 17 mRNA linker connection system

[0340]

[0341] (30) Take 0.5 μl of the original 12.5 μM adapter and dissolve it in 4.5 μl of nuclease-free water to dilute the adapter in preparation for ligation.

[0342] (31) After the second chain synthesis process is completed, transfer the PCR tube to an ice box, add 5ul of diluted adapter, and then add 45ul of adapter ligation reaction solution. Keep the PCR tube on ice and pipette up and down several times to mix. Incubate the PCR tube at 20°C for 15 minutes. Immediately perform the first post-ligation purification.

[0343] (32) After the first ligation, 0.63X magnetic bead purification was performed. 70 μL KAPA Pure Beads was added to each tube. Mix gently by pipetting up and down with a low-absorption pipette tip. The PCR tube was incubated at room temperature for 5-15 minutes to allow the DNA to bind to the magnetic beads. 80% ethanol prepared with nuclease-free water was prepared, 1200 μL / library.

[0344] (33) Place the PCR tube on a magnetic rack and incubate until the liquid is clear. Carefully aspirate and discard 175 μL of supernatant. Keep the PCR tube on the magnetic rack, add 200 μL of 80% ethanol, incubate the PCR tube on the magnetic rack at room temperature for ≥30 seconds, carefully aspirate and discard the ethanol. Keep the PCR tube on the magnetic rack, add 200 μL of 80% ethanol, incubate the PCR tube on the magnetic rack at room temperature for ≥30 seconds, carefully aspirate and discard the ethanol. Take out the PCR tube, centrifuge it briefly, and place it on the magnetic rack. Use a 10 μL pipette tip to try to aspirate all the remaining ethanol, trying not to disturb the magnetic beads.

[0345] (34) Dry the beads at room temperature for 3-5 minutes, or until all the ethanol has evaporated. Be careful not to over-dry the beads, which may result in a decrease in yield.

[0346] (35) Remove the PCR tube from the magnetic stand, completely resuspend the beads in 50 μL of EB solution, gently mix by pipetting up and down with a low-retention pipette tip, and incubate at room temperature for 2 min to elute the DNA from the beads. The eluate with the resuspended beads can be stored at 4 °C and continued for the second post-ligation purification within 24 hours.

[0347] (36) After the second ligation, 0.7X magnetic bead purification was performed. 35 μL of PEG / NaCl solution was added to each library tube using a low-retention pipette tip and mixed thoroughly by pipetting up and down. The PCR tube was incubated at room temperature for 5-15 minutes to allow the DNA to bind to the magnetic beads.

[0348] (37) Take out KAPA HiFi HotStart ReadyMix and thaw Library Amlification Primer Mix at room temperature to prepare for library amplification.

[0349] (38) Place the PCR tube on the magnetic rack after incubation and incubate until the liquid is clear. Carefully aspirate and discard 80 μL of supernatant, keep the PCR tube on the magnetic rack, and add 200 μL of 80% ethanol. Incubate the PCR tube on the magnetic rack at room temperature for ≥30 seconds and carefully aspirate and discard the ethanol. Keep the PCR tube on the magnetic rack and add 200 μL of 80% ethanol again. Incubate the PCR tube on the magnetic rack at room temperature for ≥30 seconds and carefully aspirate and discard the ethanol. Take out the PCR tube, centrifuge it briefly, and place it on the magnetic rack. Use a 10 μL pipette tip to try to aspirate all residual ethanol, trying not to disturb the magnetic beads.

[0350] (39) Dry the beads at room temperature for 3-5 minutes or until all the ethanol has evaporated. Be careful not to over-dry the beads, which may result in a decrease in yield. Remove the PCR tube from the magnetic rack and use a low-absorption pipette tip to completely resuspend the beads in 22 μL of EB solution. Incubate at room temperature for 2 minutes to elute the DNA from the beads, place the PCR tube on the magnetic rack, and incubate until the liquid is clear.

[0351] (40) Transfer 20 μL of the clear supernatant to a new PCR tube and perform library amplification.

[0352] (41) Prepare the library amplification reagents according to the following table:

[0353] Table 18 cDNA library amplification system

[0354]

[0355] (42) Add 30ul of library amplification reaction solution to the purified library in the previous step and pipette up and down several times to mix.

[0356] Perform library amplification reaction as follows:

[0357] Table 19 cDNA library amplification program

[0358]

[0359] (43) After library amplification, the library was purified immediately using 1X magnetic bead purification. 50 μL of KAPAPure Beads was added to each tube and mixed thoroughly by pipetting up and down using a low-retention pipette tip. The PCR tube was incubated at room temperature for 15 minutes to allow the DNA to bind to the magnetic beads. The PCR tube was placed on a magnetic rack and incubated until the liquid was clear. 95 μL of supernatant was carefully aspirated. Keeping the PCR tube on the magnetic rack, 200 μL of 80% ethanol was added. The PCR tube was incubated on the magnetic rack for ≥30 seconds at room temperature. The ethanol was carefully aspirated. Keeping the PCR tube on the magnetic rack, 200 μL of 80% ethanol was added again. The PCR tube was incubated on the magnetic rack for ≥30 seconds at room temperature. The ethanol was carefully aspirated. After a brief centrifugation, the PCR tube was placed on a magnetic rack and a 10 μL pipette tip was used to try to aspirate all the residual ethanol without disturbing the magnetic beads.

[0360] Dry the beads at room temperature for 3-5 minutes or until all ethanol has evaporated. Be careful to avoid over-drying the beads, which may result in reduced yield. Remove the PCR tube from the magnetic stand and completely resuspend the beads in 35 μL of EB solution. Incubate at room temperature for 2 minutes to elute the DNA from the beads.

[0361] (44) Place the PCR tube on a magnetic stand and incubate until the liquid is clear. Transfer 33 μL of the supernatant to a new 1.5 mL low-adsorption centrifuge tube for storage. Label the centrifuge tube with the library type, number, adapter, and library construction date. Use Qubit 4.0 for quantification. This is the gDNA library, which will be used in the subsequent panel cDNA probe hybridization capture experiment.

[0362] 5.3gDNA and cDNA probe working solution preparation

[0363] The A-junction Panel probe was synthesized by Integrated DNA Technologies (IDT) xGen TM Customized hybridization capture probe set. After probe synthesis, 0.75p cDNA and cDNA probe working solution need to be prepared for subsequent hybridization capture experiments.

[0364] (1) The preparation of the probe working solution is carried out in a clean bench, and the dissolving solution IDTE provided by the experimental IDT formula is used to dissolve and dilute the probe.

[0365] (2) Prepare a clean bench, 0.2 ml low-absorption flat-top tubes, sample boxes, centrifuge, vortexer, handheld centrifuge, and IDTE. Calculate the volume of IDTE required for 20 pmol / ul probe stock solution according to the instructions and the quantity of customized probe reagent.

[0366] (3) Dry powder probe was centrifuged at 12000 rpm for 5 min; corresponding volume of IDTE and votex was added for 2 min. The concentration after dilution was 20 pmol / ul. After instant centrifugation, the mixture was incubated at 4°C overnight.

[0367] (4)5. Take 30u of 20.0pmol / ul stock solution into a 200ul low adsorption tube and add 70ul IDTE. Vote x and centrifuge. The concentration after dilution is 6pmol / ul. The remaining 20pmol probe stock solution is divided into 5 small tubes, 31ul in each tube. The remaining 31ul remains in the original tube.

[0368] (5) Take 25u of 6pmol / ul stock and add 175ul of IDTE solution. Vote x and centrifuge. The concentration after dilution is 0.75pmol / ul. The remaining 6pmol stock solution is divided into 3 small tubes, 25ul in each tube.

[0369] (6) 20p stock solution is stored at -80℃, 6p stock solution is stored at -20℃. 0.75pmol / ul working solution. Dispense 100ul into 1 tube, 8.4ul into 12 tubes.

[0370] 5.4 Hybridization capture of gDNA and cDNA probe

[0371] The hybridization capture experiment of the gDNA probe of the A-junction Panel was performed in the same manner as the hybridization capture experiment of the cDNA probe and was completed using the xGenTM ​​Hybridization and Wash Kit of IDT.

[0372] (1) Take out The hybridization reagents 2X Hybridization Buffer and Hybridization Buffer Enhancer in the Hybridization and Wash Kit are melted at room temperature. Note that the 2X Hybridization Buffer must be fully melted until there are no crystals. If it cannot be melted at room temperature, it can be incubated in a 65℃ water bath, vortexing to mix until the crystals are completely melted.

[0373] (2) Turn on the power of the vacuum concentrator in advance to preheat and adjust the temperature to 60°C.

[0374] (3) Prepare a gDNA library / cDNA mixed library in a 1.5 mL low-adsorption centrifuge tube according to the following table for vacuum concentration:

[0375] Table 20 Mixed library preparation

[0376]

[0377] (4) Place the centrifuge tube in a vacuum concentrator preheated to 60°C to dry. After all the liquid has evaporated and completely dried, seal the centrifuge tube and set it aside for later use. After the centrifuge tube is sealed, it can be stored at room temperature overnight or at -20°C for a long time.

[0378] (5) Take out the A-binding Panel gDNA probe / cDNA probe working solution (0.75 μl) and thaw it on ice.

[0379] (6) Prepare the probe hybridization reaction solution according to the following table:

[0380] Table 21 Probe hybridization reaction solution

[0381]

[0382] (7) Use a pipette to mix the probe hybridization reaction solution evenly and add it to the bottom of a centrifuge tube that has been vacuum concentrated and dried. Use a pipette to gently pipette and mix 15-20 times, centrifuge briefly, and incubate at 25°C for 5-10 minutes.

[0383] (8) Vortex the hybridization reaction mixture to mix well, centrifuge briefly, transfer all 17 μL of the hybridization reaction mixture in the centrifuge tube to a new 0.2 mL PCR tube, centrifuge briefly, place in a PCR instrument, and start the following hybridization program:

[0384] Table 22 Probe hybridization reaction program

[0385]

[0386] (9) Take out 2X Hybridization Buffer, Hybridization Buffer Enhancer and other elution reagents (including 2X Bead Wash Buffer, 10X Wash BufferⅠ, 10X Wash BufferⅡ, 10X WashBufferⅢ, 10X Stringent Wash Buffer) and thaw naturally at room temperature, vortex to mix, and centrifuge briefly. Take out DynabeadsTM M-270Streptavidin Beads, vortex to mix, and equilibrate at room temperature for 30 minutes for the washing and capture steps of streptavidin affinity magnetic beads. Note that if Wash BufferⅠ cannot be dissolved, it can be incubated in a 65℃ water bath until it is completely thawed.

[0387] (10) Prepare five 1.5 mL centrifuge tubes and prepare the corresponding 1X Buffer according to the following table:

[0388] Table 23 Probe hybridization capture experiment elution buffer

[0389]

[0390] (11) Prepare another 1.5 mL centrifuge tube and label it "WBⅠ65℃". Take out 110 ul of the prepared "WBⅠ" and transfer it to the newly labeled "WBⅠ65℃" tube. Keep the remaining "WBⅠ" in the original tube for later use.

[0391] (12) Prepare the magnetic bead suspension according to the following table:

[0392] Table 24 Hybridization capture experiment magnetic bead suspension

[0393]

[0394] (13) Place "SWB 65℃" and "WBⅠ65℃" in a 65℃ metal bath and incubate for 1 hour, or at least 15 minutes.

[0395] (14) Open another PCR instrument and run the elution program as follows:

[0396] Table 25 Elution program

[0397]

[0398] (15) Prepare a 1.5 mL magnetic stand and a low-absorption 1.5 mL centrifuge tube, labeled "M270". Vortex the Dynabeads M-270 Streptavidin Beads again for 15 seconds to ensure complete mixing. Pipette 50 μL of Dynabeads M-270 Streptavidin Beads into a 1.5 mL low-absorption centrifuge tube.

[0399] (16) Add 100 μL of 1X Bead Wash Buffer to the centrifuge tube containing the magnetic beads. Mix by gently pipetting 10 times. Centrifuge briefly and place on a magnetic rack for a few minutes. When the liquid is completely clear, use a pipette to discard the supernatant. Remove the centrifuge tube from the magnetic rack.

[0400] (17) Repeat the previous step twice.

[0401] (18) Add 17 μL of magnetic bead suspension to the centrifuge tube, gently pipette to mix, and transfer all the magnetic bead suspension to a new 0.2 mL low-adsorption PCR tube. Then, place the PCR tube in a PCR instrument and incubate at 65°C for 5 min.

[0402] (19) Directly open the PCR instrument that is running the hybridization program, quickly take out the hybridization capture PCR tube, and immediately put it into the PCR instrument that is running the elution program after a brief centrifugation. Add the preheated magnetic beads to the hybridization system and use a pipette to mix. Incubate at 65°C for 45 minutes, gently mixing every 10-12 minutes to ensure that the magnetic beads are completely resuspended. Note that this part of the operation must be done quickly, as temperature changes will affect the specificity of the probe. After the sample tube is transferred, the hybridization program PCR instrument can be turned off. At the same time, take out the library amplification reagent and put it in a 4°C refrigerator to melt.

[0403] (20) Hot elution: Pipette 100 μL of 1X Wash Buffer I incubated at 65°C into a new 1.5 ml low-absorption centrifuge tube. Add all the hybridization system in the PCR tube after incubation to the 1.5 ml low-absorption centrifuge tube containing 100 μL 1X Wash Buffer I, and pipette to mix. Place the centrifuge tube on the magnetic rack for 1 min. After the liquid is completely clear, use a pipette to remove the supernatant. Remove the centrifuge tube from the magnetic rack, add 150 μL of 65°C 1X Stringent Wash Buffer, gently pipette 10 times to mix, and place in a metal bath at 65°C for 5 min.

[0404] (21) Repeat the previous step twice. The hot elution process should be performed quickly, and bubbles should be avoided during the mixing process.

[0405] (22) Elution at room temperature: Replace the vortex head with a vortex foam plate. Centrifuge the centrifuge tube for a short time and place it on a magnetic stand for 1 min. After the liquid is completely clarified, remove the supernatant and add 150 μL of room temperature 1X Wash Buffer I. Place it in the well of the vortex foam plate, adjust the vortex speed to "4-5", mix and incubate at room temperature for 2 min.

[0406] (23) After instant centrifugation, place the centrifuge tube on a magnetic rack for 1 min. After the liquid is completely clarified, remove the supernatant and add 150 μL of room temperature 1X Wash Buffer II. Place the tube in the well of a vortex foam plate and adjust the vortex speed to "4-5". Mix and incubate at room temperature for 2 min.

[0407] (24) After instant centrifugation, place the centrifuge tube on a magnetic stand for 1 min. After the liquid is completely clarified, remove the supernatant and add 150 μL of room temperature 1X Wash Buffer III. Place the tube in the well of a vortex foam plate and adjust the vortex speed to level 4-5. Mix and incubate at room temperature for 2 min.

[0408] (25) After centrifugation, place the tube on a magnetic rack for 1 min. After the liquid is completely clarified, remove the supernatant and then use a 10 μL pipette to remove a small amount of residual buffer.

[0409] (26) Remove the centrifuge tube from the magnetic stand, add 22.5 μL Nuclease Free Water, and gently pipette 10 times to ensure even mixing. Transfer all the liquid (including the magnetic beads) to a new 0.2 mL PCR tube.

[0410] (27) Take out the dissolved library amplification reagent, mix and centrifuge, and prepare the library amplification reaction system in a PCR tube placed on ice according to the following system:

[0411] Table 26 Hybridization capture library amplification system

[0412]

[0413] (28) Place the PCR tube into the PCR instrument and start the hybridization capture library amplification program:

[0414] Table 27 Hybridization capture library amplification program

[0415]

[0416] (29) Early withdrawal Beads, equilibrate at room temperature for at least 30 min, mix thoroughly, and prepare 80% ethanol.

[0417] (30) After PCR amplification, the PCR tube was removed and centrifuged briefly, then placed on a magnetic rack for 2 min. After the liquid was completely clarified, the supernatant was transferred to a new 1.5 mL low-absorption centrifuge tube.

[0418] (31) Add 60 μL of Beads, mix evenly with a pipette or vortex mixer, and incubate at 25°C for 5-10 minutes. Centrifuge the tube and place it on a magnetic rack for 5 minutes. When the liquid is completely clear, use a pipette to discard the supernatant. Slowly add 150 μL of

[0419] 80% ethanol, be careful not to disturb the magnetic beads, let it stand for 1 minute, discard the supernatant, and repeat the previous step once.

[0420] After instant centrifugation, place the centrifuge tube on a magnetic rack and use a 10μL pipette to remove a small amount of residual ethanol, being careful not to suck up the magnetic beads. Remove the centrifuge tube from the magnetic rack, open the centrifuge tube cover, and let it stand at room temperature for about 2-5 minutes until the ethanol evaporates completely.

[0421] (32) Add 20-22 μL Nuclease Free Water to the centrifuge tube, use a pipette or vortex mixer to suspend and mix the magnetic beads, and incubate in a 25°C metal bath for 2-5 min. Centrifuge the centrifuge tube briefly and place it on a magnetic stand for 1-2 min. After the liquid is completely clear, use a pipette to carefully transfer the supernatant to a new 1.5 mL low-adsorption centrifuge tube. This is the library that has completed the hybridization capture of the panel gDNA or cDNA probe for subsequent sequencing experiments.

[0422] (33) The hybridization capture library was quantified using Qubit. The library fragment distribution was detected using PerkinElmer Labchip, and the peak size of the library fragment was required to be between 300 and 500 bp.

[0423] (34)LabChip operation steps:

[0424] ① Take out the chip and reagent kit stored in the 4℃ refrigerator and warm them to room temperature for more than 30 minutes. Prepare a vacuum pump.

[0425] ② Vortex thoroughly to mix Dye Concentrate (light-proof bottle with blue cap). Gel Matrix (red cap).

[0426] ③ Prepare gel dye by adding 13 μL DNA dye concentrate to DNA Hisens / NGS3K gel matrix. Vortex to mix thoroughly.

[0427] ④ Transfer the Gel-Dye mixture to two filter membrane columns, 9200 rcf, 7.5 min. The Gel-Dye mixture can be stored at 2-8°C away from light for 3 weeks.

[0428] ⑤ During centrifugation, take out the chip for cleaning. Use a vacuum pump to suck away the ultrapure water placed in the sample well of the chip during storage.

[0429] ⑥ Use 100u ultrapure water to clean twice in 1, 3, 4, 7, and 8. When the vacuum pump sucks away the ultrapure water, pay attention to sucking out the water in the center of the bottom (except for hole No. 1, which can reduce the probability of error by properly retaining the water in the center of the bottom).

[0430] ⑦Add 50ul of Gel-Dye mixture to each of wells 3 / 7 / 8 / 10. Add 100uHiSens DNAMaker (green cap) to wells 9 and 4.

[0431] ⑧ Make sure the chip window is clean. You can use a lens cleaning cloth with or without 70% isopropanol to wipe it.

[0432] ⑨1X Ladder preparation: Use a 1.5ml EP tube to take 108ul of nuclease-free water, add 12ul of DNA ladder (yellow cap), mix well and centrifuge instantly. Transfer all liquids to the ladder tube, and do not allow any bubbles to remain.

[0433] ⑩Add 750ul of nuclease-free water to the Buffer tube.

[0434] 11. Turn on the labChip. Insert the buffer tube. Complete the pressure test (takes 3 minutes). After the test, use ultrapure water and a cotton swab to clean the needle and rubber ring.

[0435] 12Put in the ladder, put in the buffer, put in the chip, click "Prime", and complete the glue pressing (takes 20 minutes).

[0436] 13 Sample preparation. 96-well plate, total volume of each sample 15-40ul. 384-well plate, total volume of each sample 10-15ul. Concentration requirement: 200pg / ul smear strip. 20pg / fragment. It is recommended to take 2ng nucleic acid sample from the smear strip and dissolve it in nuclease-free water, with a total volume of 10ul (96-well plate).

[0437] 14 Add samples and click "Run". Confirm parameters and test plate wells. 48 samples can be tested each time. The chip can work continuously in labchip for no more than 8 hours.

[0438] 15 After the test is completed, record and save the library fragmentation information. Take out the chip. Use a vacuum pump to remove the reagents, add 120ul of ultrapure water to each well, wash twice, and add 100ul of DNA Chip Storage Buffer to each well.

[0439] 16 chips are loaded on the machine, click the "Wash" program, which takes 10 minutes.

[0440] 17 Take out the chip and store it at 4℃. For long-term storage, add water to well 1. Add ultrapure water to the chip storage box.

[0441] 18 Use a cotton swab to clean the needle and rubber ring. Turn off the LabChip. Clean the vacuum pump.

[0442] 5.5 Illumina platform next generation sequencing

[0443] The targeted enriched library of the nail node panel was sequenced using the Illumina NextSeq CN500 high-throughput sequencer.

[0444] (1) Thawing the sequencing kit: Place the reagent compartment of the sequencing kit in a water bath of room temperature deionized water, keeping the liquid level below the top of the reagent compartment. Dissolve the HT1 reagent in a 4°C refrigerator. It takes 60 minutes at this temperature. After the reagent is completely thawed (at least 60 minutes), remove the sequencing kit from the water bath and wipe the surface of the reagent compartment dry.

[0445] (2) Check whether the reagent compartment of the sequencing kit is completely thawed, invert the reagent compartment 5 times to mix the reagents, and gently tap the reagent compartment to reduce reagent bubbles.

[0446] (3) Preparation of sequencing chip: Take out the sequencing chip from the refrigerator at 2-8°C and place it at room temperature for 30 minutes. After changing into powder-free gloves, take out the sequencing chip and inspect it. Check the back of the sequencing chip to ensure that the liquid inlet and outlet are unobstructed; ensure that the gasket is installed stably and the white rectangular plastic channel is visible; ensure that the edges of the four white fixing clips are clearly beyond the edges of the black cutting board. If the black cutting board is not fixed well, gently press the black cutting board and the white frame until the cutting board is fixed in place; ensure that the four metal spring clips are flat on the black cutting board.

[0447] (4) Illumina NextSeq CN500 high-throughput sequencer was restarted after a “quick clean”.

[0448] (5) Denaturation and dilution of libraries: Prepare 0.2N NaOH and 200mM Tris-HCl, and use pH test paper to detect pH value; dilute all libraries to be sequenced to 1nM, and mix the libraries according to the data volume ratio equal to the volume ratio; take 20ul1nM mixed library, add 20ul0.2N NaOH, mix and centrifuge, and let stand for 5 minutes; add 20ul200mM Tris-HCl to neutralize; add 940ulHT1 and dilute the mixed library to 20pmol; take 91ul20pmol mixed library and add 1209ulHT1, mix and centrifuge, and the 1300ul1.4pmol mixed library is ready for sequencing.

[0449] (6) Use dust-free paper to wipe the foil-sealed reagent well No. 10, marked with "Load Library Here"; use a clean 1 ml pipette tip to pierce the foil seal and add 1.3 ml of the prepared 1.4 pmol library to well No. 10. When loading the library, avoid touching the foil seal and be careful not to create bubbles at the bottom.

[0450] (7) Set up and run the Illumina NextSeq CN500 high-throughput sequencer: Select "Sequence" from "Home" and follow the prompts to complete the subsequent operations (load the sequencing chip, empty the waste liquid collection tray, load the buffer reagent kit, etc.). After completing the pre-run check, start sequencing.

[0451] (8) Analysis of high-throughput sequencing results: FastQ files were generated by Illumina bcl2fastq software; GATK and HISAT2 software were used to align gDNA and cDNA sequences with the reference genome; Mutect2 and VarDict software were used for variant reading; Arriba and STAR-Fusion software were used to detect fusion variants; CNVkit software was used to analyze copy number variation; gene expression was normalized using the TMM method.

[0452] Example 1. Molecular typing of RAS-like thyroid nodules and non-RAS-like thyroid nodules in subjects

[0453] This study was approved by the Ethics Committee of Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine. Informed consent was obtained from all patients. Samples were obtained from postoperative tumor tissue samples of thyroid surgery performed by the Department of Thyroid Surgery of Ruijin Hospital, and 507 thyroid nodules with clear pathological diagnosis were included. The patients fully understood the sample collection and specific use of this experiment and signed the informed consent form.

[0454] Next-generation targeted sequencing of 112 genes was performed on 507 thyroid nodules with clear pathological diagnosis, and the expression and mutation of 112 genes were obtained. Thyroid nodule tissue samples were collected from patients with thyroid nodules, DNA and RNA were extracted from the tissue samples, and then the expression and mutation of genes were detected using second-generation targeted sequencing technology.

[0455] In the study of papillary thyroid carcinoma (PTC) by The Cancer Genome Atlas (TCGA), PTC was divided into BRAF-like and RAS-like according to molecular characteristics. The mutant genes of RAS-like PTC include NRAS, HRAS, KRAS, EIF1AX, PPARG fusion gene, THADA fusion gene, IGF2BP3 fusion gene, non-V600E mutated BRAF, etc. This application extends the concept from PTC to all benign and malignant thyroid nodules, and defines thyroid nodules with detected RAS-like mutant genes as "RAS-like thyroid nodules". The RAS-like mutations detected in this project involve the following mutant genes: RAS, EIF1AX, BRAF non-V600E mutation (BRAF other), PPARG fusion, THADA fusion, and IGF2BP3 fusion. According to the molecular typing results, 41 cases of RAS-like thyroid nodules and 466 cases of non-RAS-like thyroid nodules were divided.

[0456] It should be noted that in this embodiment, the sample quantity and gene expression quantity are only used for illustrative purposes and are not intended to limit the quantity.

[0457] Example 2. Gene screening and diagnostic model construction were performed using 466 cases of non-RAS-like thyroid nodules as a training set.

[0458] In order to establish a diagnostic model for distinguishing benign and malignant RAS-like thyroid nodules, it is first necessary to select important genes for diagnosis from 112 genes. This embodiment selects a series of machine learning methods to screen important genes. Variance filtering, mutual information method screening, and model method screening are performed in turn from 112 single genes to screen important genes related to benign and malignant.

[0459] Variance is an indicator that measures the degree of data variation. A small variance means that the degree of data variation is low and the information content is small. The variance calculation formula is shown in the following formula (1).

[0460]

[0461] n is the number of samples, is the mean value of X, and xi is the i-th value of X. The variances of the 112 genes were calculated respectively, and after threshold screening, 9 genes with low variance were removed, namely POU1F1, FAT1, CRYM, LHX3, TTR, SERPINA7, ALB, PROP1, and FAT4.

[0462] Next, we will perform mutual information screening. Mutual information screening is a method used to measure the correlation between genes and benign and malignant tumors. Mutual information is a method that can measure the strength of linear and nonlinear correlations. In this scenario, it is used to measure the correlation between differences in gene expression and benign and malignant tumor indicators. The calculation formula of the mutual information method is as follows:

[0463] Let μX, Y represent the joint probability density function of two random variables. The marginal density functions of these two random variables can be expressed as

[0464] μ X (x) = ∫μ X，Y (x, y)dy;μ Y (y)=∫μ X,Y (x, y)dx; (2)

[0465] Then, H(y) is the information entropy of variable y, which describes the uncertainty of y. The uncertainty of Y can be calculated as follows:

[0466] H(Y)=-∫μ Y (y)logμ Y (y)dy)

[0467] The uncertainty measure of variable Y obtained after observing another variable X is called conditional entropy, denoted as H(Y|X). Specifically, H(Y|X) can be defined as follows:

[0468] H(Y|X)=-∫μ X (x)∫μ Y (y|X=x)logμ Y (y|X=x)dy dx (4)

[0469] Then, the mutual information value between X and Y is calculated as

[0470] M(X,Y)=H(Y)-H(Y|X) (5)

[0471] Through mutual information correlation screening, 64 genes with high correlation with benign and malignant indicators of thyroid cancer were obtained, namely EML4, CTNNB1, ERG, DIO1, GAPDH, KRT7, GLIS3, BCL2, DUOXA2, IRF2BP2, MET, S100B, FGFR2, NKX2-5, BRAF, IYD, NKX2-1, IDH2, SLC16A2, THAADA, IGF2BP3, DIO2, IDH1, ACTB, PECAM1, TPO, ARAP3, AKAP9, S LC26A7, SQSTM1, KCNQ1, ENO2, ETV6, SLC5A5, PAX8, NTRK3, CALD1, STAT6, ZNF148, CHEK2, DICER1, NF1, PPARG, KRT5, CDKN2A, SLC5A8, AKT1, DIO3, TG, KCNE2, THRA, ALK, IGSF1, PGK1, CD68, TPM3, FOXE1, TLE1, EZH1, SPOP, SLC26A4, PIK3CA, TRHR and EP300.

[0472] Finally, the model screening method is used to further search for genes related to benign and malignant cancer. The model screening method used in this patent is Recursive feature elimination (RFE). RFE is a method for feature screening based on model training results, and is a method for continuously eliminating useless features through recursion. The basic execution steps of this method are as follows:

[0473] Step 1: According to the given evaluator, train on the current data set A1 to obtain the model result r1 and calculate the importance of each feature.

[0474] Step 2: Eliminate the least important feature, that is, the feature with the smallest feature importance calculation result, to obtain the feature subset A2, then train the model again to obtain the model result r2, and calculate the feature importance of the remaining features.

[0475] Step 3: If r2 is better than r1, retain A2, and further remove the least important features based on A2 to obtain feature subset A3 and train it to obtain the model result r3 at this time, as well as the importance of each feature of A3, and repeat the process continuously; conversely, if r1 is better than r2, retain A1 and stop iteration; or in the process of multiple rounds of iterations, if the modeling effect of any subset is weaker than that of the parent set, stop iteration.

[0476] After iterative screening, 9 genes were finally obtained. Figure 2It shows that the contribution of each gene in the model from high to low is TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5, and NKX2-1.

[0477] The above 9 gene names are all official names, which are unique identifiers. The CCDS number (Consensus coding sequence) is shown in Table 28. The mRNA transcription and protein translation of the corresponding gene can be downloaded from the ensembl database through the CCDS number to achieve gene sequence, mRNA sequence, and protein sequence.

[0478] Table 28. CCDS numbers of the 9 genes used to construct the diagnostic model

[0479] Gene CCDS DIO1 CCDS41339 DIO2 CCDS55934 GLIS3 CCDS43784 NKX2-1 CCDS41945 SLC26A4 CCDS5746 SLC5A5 CCDS12368 SLC5A8 CCDS9080 TG CCDS34944 TPO CCDS1643

[0480] Example 3. Construction of diagnostic model using 466 non-RAS-like thyroid nodules as training set

[0481] After determining the characteristic genes, a random forest algorithm was used to construct a diagnostic model for the differential diagnosis of benign and malignant RAS thyroid nodules based on the expression levels of these 9 important genes.

[0482] Random forest refers to the use of multiple trees to train samples to obtain a classifier containing multiple decision trees. The specific process of model establishment is described as follows. There are 466 non-RAS-like thyroid nodules in the training set, 50 benign nodules, and 416 malignant nodules. Therefore, 466 samples are randomly selected with replacement (that is, one sample is randomly selected each time, and then returned to continue selection).

[0483] Then, the 466 samples are selected to train a decision tree as the samples at the root node of the decision tree. When each node of the decision tree needs to be split, m (m<=5) genes are randomly selected from 9 genes, where m is a positive integer. Each decision tree selects genes through the Gini index, and the gene with the smallest Gini index will be used as the split condition of the node.

[0484] 18 decision trees are constructed according to the above steps through the bootstrap resampling technology. The maximum depth of each decision tree is 3 layers, thus forming a random forest. When a data set is input into the random forest model, the 18 decision trees in the random forest will diagnose the data set respectively. The random forest will comprehensively evaluate the decision results of all decision trees and use the majority principle to vote and output the final diagnosis result. For example, for a certain data set, 10 of the 18 decision trees diagnose the sample as malignant and 8 diagnose the sample as benign. Then the random forest will eventually output the result of diagnosing the sample as malignant.

[0485] Example 4. Validation of the diagnostic model

[0486] The diagnostic model was used to determine the benign or malignant nature of 41 RAS-like thyroid nodules.

[0487] Figure 3 Table 29 shows the results of predicting the benign and malignant nature of thyroid nodules using the diagnostic model constructed by the present invention in 41 cases of RAS-like thyroid nodules.

[0488]

[0489]

[0490] Taking pathological diagnosis as the gold standard, the specificity, sensitivity, positive predictive rate, negative predictive rate and accuracy of the diagnostic method were calculated to evaluate the diagnostic efficacy.

[0491] Sensitivity: the positive consistency rate, which indicates the consistency rate of the positive results of the evaluated method with those of the previous method when the results are assumed to be true and reliable. In this study, "previous method" refers to pathological diagnosis.

[0492] Specificity: It is the negative consistency rate, which indicates the consistency rate of the negative results of the evaluation methodology with the previous methods when the negative results are assumed to be true and reliable results.

[0493] Positive Predictive Value (PPV) is the ratio of the results of the evaluation method that are positive when the previous method is also positive.

[0494] Negative Predictive Value (NPV) is the ratio of negative results of the evaluation method to those that were also negative in the previous method.

[0495] Accuracy (ACC): A statistical analysis of positive and negative results to assess the overall consistency of a methodology compared with previous methodologies.

[0496] As shown in Table 30, the diagnostic model had a sensitivity of 90.00%, a specificity of 70.97%, a positive predictive rate of 50%, a negative predictive rate of 95.65%, and an accuracy of 75.61% (59.7-87.64%).

[0497] Table 30. Diagnostic efficacy of the RAS-like model for distinguishing benign and malignant thyroid nodules

[0498]

[0499]

[0500] Equivalents and scope

[0501] It is to be understood that the present invention is not limited to any or all of the specific embodiments explicitly described herein, and as such may, of course, vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting.

[0502] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the invention belongs. Although any methods and materials similar or equivalent to those described herein can also be used in the practice or testing of the present invention, the preferred methods and materials are currently described.

[0503] All publications and patents cited in this specification are cited to disclose and describe methods and / or materials related to the cited publications. All such publications and patents are incorporated herein by reference to the same extent as each individual publication or patent specifically and individually indicates the same extent of incorporation by reference. Such incorporation by reference is explicitly limited to the methods and / or materials described in the cited publications and patents, and is not extended to any dictionary definitions in the cited publications and patents (that is, any dictionary definitions in the cited publications and patents that are not explicitly repeated in the present invention should not be considered as such and should not be understood as defining any term appearing in the attached claims). If there is a conflict between any incorporated reference and the present invention, the present invention shall prevail. In addition, any specific embodiment of the present invention falling within the scope of the prior art may be explicitly excluded from any one or more claims. Because such embodiments are considered to be known to those of ordinary skill in the art, they may also be excluded even if the exclusion is not explicitly set forth herein. Any specific embodiment of the present invention may be excluded from any claim for any reason, whether or not it is related to the existence of the prior art.

[0504] The citation of any publication is for its disclosure prior to the filing date and should not be construed as an admission that the present invention is not entitled to antedate such publication by virtue of prior disclosure. Further, the dates of publication provided may be different from the actual publication dates and may need to be independently confirmed.

[0505] As will be apparent to those skilled in the art after reading this disclosure, each individual embodiment described and illustrated herein has discrete components and features that can be easily separated or combined with the features of any of the other several embodiments without departing from the scope or spirit of the invention. Any enumerated method can be performed in the order of the enumerated events or in any other order that is logically possible.

[0506] In the claims, articles such as "a", "an", and "the" may mean one or more, unless otherwise indicated or obvious from the context. Regardless of the gender pronouns used herein (e.g., masculine, feminine, neuter, other, etc., etc.), the pronouns should be interpreted as gender neutral (i.e., interpreted as referring to all genders equally) regardless of the implied gender, unless the context clearly indicates or requires otherwise. Unless the context clearly indicates or requires otherwise, no matter where used in this document, words used in the singular include the plural form, and words used in the plural form include the singular form. If one, more than one, or all of the group members are present in, used in, or otherwise related to a given product or process, a claim or description including "or" between one or more members of the group is deemed to be satisfied, unless otherwise indicated or obvious from the context. The present invention includes embodiments in which exactly one member of the group is present in, used in, or otherwise related to a given product or process. The present invention includes embodiments in which more than one or all of the group members are present in, used in, or otherwise related to a given product or process.

[0507] In addition, the present invention encompasses all variations, combinations and arrangements, wherein one or more limitations, elements, clauses and descriptive terms of one or more listed claims are introduced into another claim. For example, any claim that is subordinate to another claim may be modified to include one or more limitations found in any other claim that is subordinate to the same basic claim. In the case where an element is presented in a list form (e.g., in a Markush group format), each subgroup of the element is also disclosed, and any element may be deleted from the group. It should be understood that, in general, in the case where the disclosed or disclosed aspects are referred to as comprising specific elements and / or features, certain embodiments of the present invention or aspects of the present invention are composed of such elements and / or features or are substantially composed of such elements and / or features. For simplicity, those embodiments are not specifically described herein. It should also be noted that the terms "comprising" and "containing" are intended to be open and allow the inclusion of additional elements or steps. In the case of a given range, the endpoints are included within this range unless otherwise stated. In addition, unless otherwise indicated or apparent from the context and understanding of one of ordinary skill in the art, values ​​expressed as ranges may adopt any specific value or sub-range within the range described in different embodiments of the present invention, to the tenth of the unit of the lower limit of the range, unless the context clearly dictates otherwise.

[0508] Those skilled in the art will recognize or be able to determine many equivalents of the specific embodiments described herein using only routine experimentation. The scope of the present embodiments described herein is not intended to be limited to the above description, but as set forth in the appended claims. Those of ordinary skill in the art will appreciate that various changes and modifications may be made to this description without departing from the spirit or scope of the present invention as defined in the appended claims.

[0509] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the embodiments described herein.The scope of the present invention is not intended to be limited to the above description, but rather is set forth in the appended claims.

[0510] Articles such as "a", "an", and "the" may mean one or more, unless otherwise indicated or obvious from the context. Claims or descriptions that include "or" between two or more members of the group are deemed to be satisfied if one, more than one, or all of the group members are present, unless otherwise indicated or obvious from the context. Disclosure of a group that includes "or" between two or more group members provides embodiments where exactly one member of the group is present, embodiments where more than one member of the group is present, and embodiments where all members of the group are present. For the sake of brevity, those embodiments are not individually set forth herein, but it is understood that each of these embodiments is provided herein and may be specifically claimed or disclaimed.

[0511] It should be understood that the present invention covers all variations, combinations and arrangements in which one or more limitations, elements, clauses or descriptive terms from one or more claims or from one or more relevant parts of the specification are introduced into another claim. For example, any claim that is dependent on another claim can be modified to include one or more limitations found in any other claim that is dependent on the same basic claim. In addition, where the claims recite a composition, it should be understood that it includes methods of making or using the composition according to any method disclosed herein or according to methods known in the art (if any), unless otherwise indicated or unless it is obvious to a person of ordinary skill in the art that a contradiction or inconsistency would occur.

[0512] In the case where an element is presented in a list form, for example, in a Markush group format, it should be understood that each possible subgroup of an element is also disclosed, and any element or element subgroup can be removed from the group. It should also be noted that the term "comprising" is intended to be open and allows to include additional elements or steps. It should be understood that, usually when an embodiment, product or method is referred to as comprising a specific element, feature or step, an embodiment, product or method consisting of or substantially consisting of such elements, features or steps is also provided. For the sake of brevity, those embodiments are not separately set forth in this article, but it should be understood that each of these embodiments is provided in this article and can specifically claim protection or abandonment.

[0513] Where ranges are given, endpoints are also included. In addition, it should be understood that, unless otherwise indicated or apparent from the context and / or understanding of a person of ordinary skill in the art, in some embodiments, values ​​expressed as ranges may take any particular value within the ranges described in some embodiments, to one-tenth of the units of the lower limit of the range, unless otherwise clearly specified in the context. For the sake of brevity, the values ​​within each range are not separately set forth herein, but it should be understood that each of these values ​​is provided herein and may be specifically claimed or waived. It should also be understood that, unless otherwise indicated or apparent from the context and / or understanding of a person of ordinary skill in the art, values ​​expressed as ranges may take any sub-range within a given range, wherein the endpoints of the sub-ranges are expressed with the same precision as one-tenth of the units of the lower limit of the range.

[0514] In addition, it should be understood that any specific embodiment of the present invention may be explicitly excluded from any one or more claims. Where a range is given, any value within the range may be explicitly excluded from any one or more claims. Any embodiment, element, feature, application or aspect of the composition and / or method of the present invention may be excluded from any one or more claims. For the sake of brevity, all embodiments excluding one or more elements, features, objects or aspects are not explicitly set forth herein.

[0515] The phrase "and / or" as used in the specification and claims should be understood to mean "one or both" of the elements so combined, i.e., elements that are present in conjunction in some cases and separately in other cases. Multiple elements listed with "and / or" should be interpreted in the same manner, i.e., "one or more" of the elements so combined. In addition to the elements specifically identified by the "and / or" clause, other elements may optionally be present, whether related or unrelated to those specifically identified elements. Thus, as a non-limiting example, when used in conjunction with an open-ended statement such as "comprising," a reference to "A and / or B" may refer to only A (optionally including elements other than B) in one embodiment; may refer to only B (optionally including elements other than A) in another embodiment; may refer to both A and B (optionally including other elements) in yet another embodiment; and so on.

[0516] As used herein in the specification and claims, "or" should be understood to have the same meaning as "and / or" as defined above. For example, when separating items in a list, "or" or "and / or" should be interpreted as inclusive, i.e., including at least one element, but also including more than one element of a plurality of elements or list elements, and optionally other unlisted items. Only terms that clearly indicate the opposite, such as "only one" or "exactly one", or when "consisting of..." is used in a claim, it means that exactly one element of a plurality of elements or list elements is included. Generally, the term "or" as used herein should only be interpreted as indicating exclusive alternatives (i.e., "one or the other, but not both"), provided that there are exclusive terms, such as "any one", "one of them", "only one" or "exactly one of them". "Substantially consisting of...", when used in a claim, should have the ordinary meaning in the field of patent law.

Claims

1. Use of a reagent for detecting the expression level of a gene in a sample of a subject in the preparation of a kit for diagnosing or predicting thyroid cancer in the subject or identifying the benign or malignant nature of a thyroid nodule in the subject, wherein the gene comprises one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

2. The method according to claim 1, wherein the expression level of the gene is the level of mRNA transcribed from the gene.

3. The use according to claim 1, wherein the expression level of the gene is the level of protein translated from the gene.

4. The method according to any one of claims 1 to 3, wherein the sample is a tumor tissue, preferably, the tumor tissue is a liquid nitrogen frozen tissue sample, an FNA (fine needle aspiration) sample and / or a FFPE sample.

5. The use according to any one of claims 1 to 3, wherein the sample is blood.

6. The use according to any one of claims 1 to 5, wherein the subject is a human.

7. The use according to any one of claims 1 to 6, wherein the gene comprises one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG and SLC5A5.

8. The use according to any one of claims 1 to 6, wherein the gene comprises one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2 and TG.

9. The use according to any one of claims 1 to 6, wherein the gene comprises one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1 and DIO2.

10. The use according to any one of claims 1 to 6, wherein the gene comprises one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3 and DIO1.

11. The use according to any one of claims 1 to 6, wherein the gene comprises one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4 and GLIS3.

12. The use according to any one of claims 1-6, wherein the gene comprises one or more genes selected from the group consisting of: TPO, SLC5A8, and SLC26A4.

13. The use according to claim 2, wherein: The reagent comprises a primer molecule that hybridizes with the mRNA sequence or its fragment or its transformed variant, and the primer molecule can reverse transcribe and amplify cDNA. Preferably, the reagent also comprises a probe molecule that hybridizes with the cDNA sequence or its fragment or its transformed variant.

14. The use according to any one of claims 1 to 13, wherein the diagnosis comprises: A score is obtained by calculation, and the thyroid cancer of the subject is diagnosed or predicted or the benign or malignant thyroid nodule of the subject is identified according to the score.

15. The use according to claim 14, wherein the calculation is performed by a random forest model.

16. A kit for diagnosing thyroid cancer in the subject or identifying benign or malignant thyroid nodules in the subject, the kit comprising reagents for detecting the expression levels of genes in a sample from the subject, wherein the genes comprise one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

17. The kit according to claim 17, comprising a chip, and / or primer molecules that hybridize with mRNA sequences transcribed from the gene or fragments thereof or transformed variants thereof, the primer molecules being capable of reverse transcription and amplification of cDNA, and / or probe molecules that hybridize with the cDNA sequences or fragments thereof or transformed variants thereof.

18. Use of gene expression levels in constructing a computational model for diagnosing or predicting thyroid cancer in the subject or identifying benign or malignant thyroid nodules in the subject, wherein the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1.

19. The use according to claim 18, wherein The input variable of the computational model is the expression level of the gene, preferably the level of mRNA transcribed from the gene, and more preferably the level of mRNA transcribed from the gene in tumor tissue.

20. The use according to claim 19, wherein The calculation is to use the constructed thyroid cancer diagnostic model to obtain a diagnostic result model indicating whether the subject's thyroid nodule is benign or malignant.

21. The use according to claim 20, wherein The thyroid cancer diagnosis model is a random forest model.

22. A device for diagnosing or predicting thyroid cancer or distinguishing benign and malignant thyroid nodules, the device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the following steps are implemented: (1) obtaining the expression level of genes in a sample of the subject, wherein the genes include one or more genes selected from the group consisting of: TPO, SLC5A8, SLC26A4, GLIS3, DIO1, DIO2, TG, SLC5A5 and NKX2-1, (2) inputting the expression level of the gene into a thyroid cancer diagnosis model based on a random forest algorithm, and (3) Diagnose thyroid cancer or identify thyroid nodules as benign or malignant based on the output results of the thyroid cancer diagnostic model.

23. The device according to claim 22, wherein step (1) comprises detecting the expression level of the gene in the sample, preferably comprises detecting the level of mRNA transcribed from the gene in the sample, and more preferably comprises detecting the level of mRNA transcribed from the gene in tumor tissue, and the detection includes but is not limited to: RNA extraction, reverse transcription, PCR, and DNA sequencing.

24. A device for diagnosing or predicting thyroid cancer or distinguishing between benign and malignant thyroid nodules, comprising: A data acquisition unit is configured to acquire sample data including tissue samples of multiple normal persons and sample data including tissue samples of multiple thyroid nodule patients as a first training data set, wherein the first training data set includes a first number of sample data, and each of the sample data includes abundance data of gene expression of a second number of types; Diagnostic model construction unit, including: a feature screening unit, configured to perform dimensionality reduction processing on the first training data set using a lasso regression algorithm to obtain a second training data set, wherein the second training data includes a first number of dimensionality reduction sample data, and each of the dimensionality reduction sample data includes abundance data of gene expression of a third number of types screened from the second number of types of genes, preferably, important genes related to benign and malignant are screened by variance filtering, mutual information method screening, and model method screening, and preferably, the model screening method is Recursive feature elimination (RFE); A model construction unit is configured to utilize a random forest algorithm to train multiple decision trees using the second training data set to construct a random forest model as a thyroid cancer diagnosis model; and a diagnosis unit is configured to use the constructed thyroid cancer diagnosis model to obtain a diagnosis result indicating whether the subject's thyroid nodule is benign or malignant.

25. A computer storage medium storing instructions, which, when executed by one or more computers, causes the one or more computers to implement the apparatus of any one of claims 22-24.

Images