Composite biomarkers for cancer immunotherapy

Abstract

To provide a system and method for determining a composite biomarker based on genomic metric and transcriptomic metric from a biological sample.SOLUTION: Provided is a method for generating a composite biomarker that identifies a predicted level of responsiveness of a subject to a particular type of an immunotherapy treatment. The method can include generating genomic metrics that represent one or more characteristics corresponding to one or more DNA sequences. The method can also include generating transcriptomic metrics that represent one or more characteristics corresponding to a set of peptides that are translated from a corresponding RNA sequence of the one or more RNA sequences. The method can also include generating a composite biomarker score derived from the set of genomic metrics and the set of transcriptomic metrics. The method can also include determining, based on the composite biomarker score, a predicted level of responsiveness of the subject to a particular type of an immunotherapy treatment.SELECTED DRAWING: Figure 1

Description

FIELD OF THEINVENTION

[0001] The present disclosure relates generally to systems and methods for determining composite biomarkers based on genomic and transcriptomic indices derived from a biological sample. More particularly, but not exclusively, the present disclosure relates to determining a composite biomarker score based on genomic and transcriptomic indices that identifies a predicted level of a subject's responsiveness to a particular type of immunotherapy treatment. [Background technology]

[0002] Related Patent Cross-References This application claims priority to U.S. Provisional Patent Application No. 63 / 017,542, filed April 29, 2020, and U.S. Provisional Patent Application No. 63 / 040,943, filed June 18, 2020, each of which is incorporated by reference in its entirety for all purposes.

[0003] Immunotherapy is used to treat many cancers and autoimmune diseases.Immune checkpoint blockade therapy is known to be an effective form of cancer treatment for various malignant tumors, but diagnostic biomarkers that consistently predict the responsiveness of subjects to these therapies have yet to be found.Given the highly diverse and complex nature of immune system resistance to immunotherapy and the potential toxicity associated with treatment, it may be difficult to accurately predict the therapeutic responsiveness to a particular immunotherapy.

[0004] Immunogenomics has emerged as a technique that can determine the therapeutic efficacy of immunotherapy. This technique can lead to the determination of effective cancer treatment and may contribute to the discovery of several new therapies, diagnostics, and processes. For example, immunogenomics can be used to identify neoantigens, which can contribute to the development of precise cancer treatments and diagnostics. Furthermore, genomic data such as variant calling can provide insights into the complex immune system response and resistance to cancer immunotherapy. However, traditional techniques using cancer endpoints for targeted diagnosis provide only limited data and may be unreliable for the development of integrated composite biomarkers. Summary of the Invention

[0005] In some embodiments, methods and systems are provided for determining a composite biomarker score that identifies a predicted level of a subject's responsiveness to a particular type of immunotherapy treatment. The immunogenomic analysis system processes a subject's biological sample to obtain genomic and transcriptomic data. In some cases, the biological sample includes one or more cancer cells. The genomic data can identify one or more DNA sequences in the biological sample, where whole exome sequencing can be performed to identify one or more DNA sequences. The transcriptomic data can identify one or more RNA sequences in the biological sample, where transcriptomic sequencing can be used to identify one or more RNA sequences. Additionally or alternatively, the genomic and transcriptomic data can be generated from a sample pair that includes the subject's biological sample and a reference biological sample, where the reference biological sample does not include one or more cancer cells.

[0006] In an immunogenomics analysis system, the genomic data is processed to generate a set of genomic indices. Each set of genomic indices can represent one or more characteristics corresponding to the corresponding DNA sequence as one or more DNA sequences. In some cases, the set of genomic indices includes (i) a quantitative or classifying indices representing one or more characteristics for each of one or more somatic mutations in the one or more DNA sequences; (ii) a classifying indices indicating whether loss of heterozygosity has occurred in at least one human leukocyte antigen (HLA) gene of the biological sample; and (iii) a quantitative or classifying indices representing a predicted tumor mutation load. For HLA loss of heterozygosity, the corresponding classifying indices can be generated by fitting the genomic data to an HLA deletion discrimination machine learning model.

[0007] The immunogenomics analysis system processes the transcriptome data to generate a set of transcriptome indices, each of which can represent one or more features corresponding to a set of peptides translated from a corresponding RNA sequence of the one or more RNA sequences. In some cases, the set of transcriptome indices includes: (i) a quantitative index or classifier representing a predicted neoantigen load of the biological sample; (ii) a quantitative index or classifier representing one or more features of each of the one or more candidate neoantigens detected in the biological sample; (iii) a quantitative index or classifier representing one or more features of each of the one or more HLA proteins for which loss of cell surface presentation is detected; (iv) a quantitative index or classifier representing one or more features corresponding to HLA genes encoding the one or more HLA proteins for which loss of cell surface presentation is detected; (v) a quantitative index or classifier representing an expression level of a sequence corresponding to an immune cell; and (vi) a quantitative index or classifier representing an expression level of one or more T cell receptors detected in the biological sample. For HLA proteins where loss of cell surface presentation is detected, corresponding indices can be generated by fitting genomic and transcriptomic data to neoantigen presentation predictive machine learning models.

[0008] The immunogenomic analysis system generates a composite biomarker score derived from the set of genomic indices and the set of transcriptomic indices, and determines a predicted level of a subject's responsiveness to a particular type of immunotherapy treatment based on the composite biomarker score. In some cases, the immunogenomic analysis system generates the composite biomarker score by (i) weighting each genomic indices in the set of genomic indices with a weight value determined based on a corresponding transcriptomic indices in the set of transcriptomic indices, and (ii) using the weighted genomic indices to generate the composite biomarker score.

[0009] The immunogenomic analysis system outputs a result corresponding to a predicted level of responsiveness of the subject. The result may be a report that identifies (i) a recommendation of a particular treatment; (ii) a recommendation to administer the particular treatment to the human subject; and / or (iii) a recommendation not to administer the particular treatment to the human subject based on the predicted level of responsiveness of the subject to the particular treatment. In some embodiments, the recommended treatment is administered to the human subject.

[0010] In some embodiments, a computer program product is provided that is tangibly embodied in a non-transitory machine-readable storage medium and includes instructions configured to cause one or more data processors to perform some or all of one or more of the methods disclosed herein.

[0011] Some embodiments of the disclosure include a system including one or more data processors. In some embodiments, the system includes a non-transitory computer-readable storage medium including instructions that, when executed on the one or more data processors, cause the one or more data processors to perform some or all of one or more methods and / or some or all of one or more processes disclosed herein. Some embodiments of the disclosure include a computer program product tangibly embodied in a non-transitory machine-readable storage medium, including instructions configured to cause the one or more data processors to perform some or all of one or more methods and / or some or all of one or more processes disclosed herein.

[0012] The terms and expressions used are used as terms of description and not for purposes of limitation, and the use of such terms and expressions is not intended to exclude the features shown and described or equivalents of any portion thereof, knowing that various modifications are possible within the scope of the claims of the present invention. It is therefore to be understood that although the claimed invention has been specifically disclosed by embodiments and optional features, those skilled in the art may make modifications and changes to the spirit of the disclosure herein, and such modifications and changes are considered to be within the scope of the present invention as defined by the appended claims.

[0013] The features, embodiments, and advantages of the present disclosure will be more clearly understood from the following detailed description when read in conjunction with the following drawings: The patent or application file contains drawings executed in at least one color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee. [Brief description of the drawings]

[0014] [Figure 1] FIG. 1 shows an example schematic for generating genomic and transcriptomic data from a biological sample, according to some embodiments. [Figure 2A]2A-B show statistical data corresponding to oncogenic alterations in genomic and transcriptomic data corresponding to subjects in a clinical population. [Figure 2B] 2A-B show statistical data corresponding to oncogenic alterations in genomic and transcriptomic data corresponding to subjects in a clinical population. [Figure 3A] 3A-C show statistical data corresponding to the transcriptome index identifying differentially expressed genes associated with immune system response. [Figure 3B] 3A-C show statistical data corresponding to the transcriptome index identifying differentially expressed genes associated with immune system response. [Figure 3C] 3A-C show statistical data corresponding to the transcriptome index identifying differentially expressed genes associated with immune system response. [Figure 4] FIG. 4 shows the statistical data corresponding to the normalized enrichment scores for each differentially regulated immune pathway. [Figure 5A] 5A-C show statistical data corresponding to the transcriptomic signature identifying the expression levels of T cell receptors. [Figure 5B] 5A-C show statistical data corresponding to the transcriptomic signature identifying the expression levels of T cell receptors. [Figure 5C] 5A-C show statistical data corresponding to the transcriptomic signature identifying the expression levels of T cell receptors. [Figure 6] FIG. 6 shows a set of box plots identifying a comparison of enrichment scores between a first group of responding subjects and a second group of non-responding subjects. [Figure 7A] 7A-B show statistical data corresponding to the transcriptomic signature identifying neoantigen burden across different genes and disease sites. [Figure 7B] 7A-B show statistical data corresponding to the transcriptomic signature identifying neoantigen burden across different genes and disease sites. [Figure 8A]8A-F show statistical data identifying neoantigen burden scores across subjects, which can predict responsiveness of subjects treated with immunotherapy. [Figure 8B] 8A-F show statistical data identifying neoantigen burden scores across subjects, which can predict responsiveness of subjects treated with immunotherapy. [Figure 8C] 8A-F show statistical data identifying neoantigen burden scores across subjects, which can predict responsiveness of subjects treated with immunotherapy. [Figure 8D] 8A-F show statistical data identifying neoantigen burden scores across subjects, which can predict responsiveness of subjects treated with immunotherapy. [Figure 8E] 8A-F show statistical data identifying neoantigen burden scores across subjects, which can predict responsiveness of subjects treated with immunotherapy. [Figure 8F] 8A-F show statistical data identifying neoantigen burden scores across subjects, which can predict responsiveness of subjects treated with immunotherapy. [Figure 9A] 9A-F show statistical data identifying one or more characteristics related to the mutations present in each subject sample in the discovery cohort. [Figure 9B] 9A-F show statistical data identifying one or more characteristics related to the mutations present in each subject sample in the discovery cohort. [Figure 9C] 9A-F show statistical data identifying one or more characteristics related to the mutations present in each subject sample in the discovery cohort. [Figure 9D] 9A-F show statistical data identifying one or more characteristics related to the mutations present in each subject sample in the discovery cohort. [Figure 9E] 9A-F show statistical data identifying one or more characteristics related to the mutations present in each subject sample in the discovery cohort. [Figure 9F] 9A-F show statistical data identifying one or more characteristics related to the mutations present in each subject sample in the discovery cohort. [Figure 10] FIG. 10 shows multiple sets of box plots identifying tumor mutation burden across different driver mutations, disease sites, and subject groups. [Figure 11A] 11A-D show statistical data identifying a composite biomarker score across subjects that demonstrates improved performance in predicting responsiveness of subjects treated with immunotherapy. [Figure 11B] 11A-D show statistical data identifying a composite biomarker score across subjects that demonstrates improved performance in predicting responsiveness of subjects treated with immunotherapy. [Figure 11C] 11A-D show statistical data identifying a composite biomarker score across subjects that demonstrates improved performance in predicting responsiveness of subjects treated with immunotherapy. [Figure 11D] 11A-D show statistical data identifying a composite biomarker score across subjects that demonstrates improved performance in predicting responsiveness of subjects treated with immunotherapy. [Figure 12A] 12A-B show statistical data identifying a composite biomarker score across a range of subjects that demonstrates improved performance in predicting progression-free survival and overall survival of subjects in a population. [Figure 12B] 12A-B show statistical data identifying a composite biomarker score across a range of subjects that demonstrates improved performance in predicting progression-free survival and overall survival of subjects in a population. [Figure 13A] 13A-B show statistical data identifying somatic mutations to HLA genes that may contribute to a decreased probability of neoantigen presentation. [Figure 13B]13A-B show statistical data identifying somatic mutations to HLA genes that may contribute to a decreased probability of neoantigen presentation. [Figure 14A] 14A-B show examples of sets of endpoints specifying comparison of HLA sequences between normal and matched tumor samples of a particular subject. [Figure 14B] 14A-B show examples of sets of endpoints specifying comparison of HLA sequences between normal and matched tumor samples of a particular subject. [Figure 15] FIG. 15 includes a flow diagram illustrating an example of a method for generating a composite biomarker score, according to some embodiments. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] I. Overview As mentioned above, the efficacy of checkpoint blockade therapy may depend on various biological factors, including the complex interactions between the tumor, the corresponding tumor microenvironment, and the corresponding immune system. Numerous biomarkers have been considered to identify the immune system's response to immunotherapy, including PD-L1 expression, interferon (IFN)-γ-based signature, tumor mutation load, mismatch repair deficiency, genetic alterations including changes within the antigen presentation machinery, loss of heterozygosity of HLA, and diversity of T cell repertoire.

[0016] As indicated by the variety of biological factors that can affect the immune system response to immune checkpoint blockade therapy, there is an expanding search for integrative biomarkers that incorporate various biological factors and can accurately predict the immune system response to immunotherapy. For example, traditional techniques have combined information corresponding to the immunogenicity and neoantigen clonal structure of a sample to predict the immune system response to immune checkpoint blockade. Results generated by these traditional techniques have attempted to determine the prognosis of subjects suffering from melanoma, lung cancer, and kidney cancer. Although these traditional techniques have yielded some positive results, the traditional techniques are still inadequate in generating data that can consistently and accurately predict the immune system response. This challenge may be due to the complex mechanisms that drive the immune response against tumors. Furthermore, these traditional techniques require a large amount of sample from the subject, which makes them invasive and may be difficult to obtain in some situations (e.g., when the subject is elderly or young, or when the subject is pregnant).

[0017] To address at least the above-mentioned shortcomings of conventional systems, the technology of the present invention can be used to determine a composite biomarker score that identifies a predicted level of a subject's responsiveness to a particular type of immunotherapy treatment. The immunogenomics analysis system processes a subject's biological sample to obtain genomic and transcriptomic data. In some cases, the biological sample includes one or more cancer cells. The genomic data can identify one or more DNA sequences in the biological sample, where whole exome sequencing can be performed to identify one or more DNA sequences. The transcriptomic data can identify one or more RNA sequences in the biological sample, where transcriptomic sequencing can be used to identify one or more RNA sequences. Additionally or alternatively, the genomic and transcriptomic data can be generated from a sample pair that includes the subject's biological sample and a reference biological sample, where the reference biological sample does not include one or more cancer cells.

[0018] The immunogenomics analysis system processes the genomic data to generate a set of genomic indices. Each of the set of genomic indices can represent one or more characteristics corresponding to the corresponding DNA sequence as one or more DNA sequences. In some cases, the set of genomic indices includes (i) a quantitative or classifying indices representing one or more characteristics for each of one or more somatic mutations in the one or more DNA sequences; (ii) a classifying indices indicating whether loss of heterozygosity has occurred within at least one human leukocyte antigen (HLA) gene of the biological sample; and (iii) a quantitative or classifying indices representing a predicted tumor mutation load. For HLA loss of heterozygosity, the corresponding classifying indices can be generated by fitting the genomic data to an HLA deletion discrimination machine learning model.

[0019] The immunogenomics analysis system processes the transcriptome data to generate a set of transcriptome indices. Each set of transcriptome indices can represent one or more characteristics corresponding to a set of peptides translated from a corresponding RNA sequence of the one or more RNA sequences. In some cases, the set of transcriptome indices includes: (i) a quantitative index or classifier representing a predicted neoantigen load of the biological sample; (ii) a quantitative index or classifier representing one or more characteristics of each of one or more candidate neoantigens detected in the biological sample; (iii) a quantitative index or classifier representing one or more characteristics of each of one or more HLA proteins for which loss of cell surface presentation is detected; (iv) a quantitative index or classifier representing one or more characteristics corresponding to HLA genes encoding one or more HLA proteins for which loss of cell surface presentation is detected; (v) a quantitative index or classifier representing an expression level of a sequence corresponding to an immune cell; and (vi) a quantitative index or classifier representing an expression level of one or more T cell receptors detected in the biological sample. For HLA proteins where loss of cell surface presentation is detected, corresponding indices can be generated by fitting genomic and transcriptomic data to neoantigen presentation predictive machine learning models.

[0020] The immunogenomic analysis system generates a composite biomarker score derived from the set of genomic indices and the set of transcriptomic indices, and determines a predicted level of a subject's responsiveness to a particular type of immunotherapy treatment based on the composite biomarker score. In some cases, the immunogenomic analysis system generates the composite biomarker score by (i) weighting each genomic indices in the set of genomic indices with a weight value determined based on a corresponding transcriptomic indices in the set of transcriptomic indices, and (ii) using the weighted genomic indices to generate the composite biomarker score.

[0021] The immunogenomic analysis system outputs a result corresponding to a predicted level of responsiveness of the subject. The result may be a report that identifies (i) a recommendation of a particular treatment; (ii) a recommendation to administer the particular treatment to the human subject; and / or (iii) a recommendation not to administer the particular treatment to the human subject based on the predicted level of responsiveness of the subject to the particular treatment. In some embodiments, the recommended treatment is administered to the human subject.

[0022] Thus, embodiments of the present disclosure provide technical advantages over conventional techniques by generating a composite biomarker score based on a validated and enhanced exome- and transcriptome-based tumor identification platform. In particular, the composite biomarker score can be determined from indices characteristic of various tumor and immune-related molecular mechanisms, and the amount of biological sample used to generate the indices can be minimized. Such techniques may improve the accuracy of diagnosis, prognosis, and / or treatment recommendations for corresponding subjects without the need for invasive procedures to obtain large amounts of biological samples. Thus, embodiments of the present disclosure provide a composite immunogenomics framework for accurately predicting responsiveness to immunotherapy treatment by identifying biological mechanisms that promote responsiveness and resistance to the immunotherapy.

[0023] While various embodiments of the disclosed invention have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Numerous variations, modifications, and substitutions may occur to those skilled in the art without departing from the invention. It will be appreciated that various alternative techniques to the embodiments of the invention described herein may be used in implementing any one of the inventions described herein.

[0024] II. Definition While various embodiments of the present invention have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Numerous variations, changes, and substitutions may occur to those skilled in the art without departing from the invention. It will be appreciated that various alternative techniques may be used to the embodiments of the invention described herein.

[0025] The term "cancer" or "malignant tumor" as used herein generally refers to a collection of related diseases in which the body's cells divide without stopping and spread into surrounding tissues. Cancer can arise almost anywhere in the body and grow when the regular process of removing and replacing aged, abnormal, or damaged cells is interrupted, these cells survive when they should die, or new cells are formed when they are not needed. These cells divide without stopping and can spread and invade both nearby and distant tissues from their point of origin.

[0026] The term "neoantigen" as used herein generally refers to a newly formed antigen that has not been previously recognized by the immune system. Neoantigens may result from changes in tumor proteins that form as a result of tumor mutations. Neoantigens may constitute a subset of somatic mutations that can be loaded onto the surface of MHC class I and class II molecules and presented to T cells. These neoantigens may be recognized by the immune system as endogenous tumor-specific (non-self) targets.

[0027] The term "tumor microenvironment" as used herein refers to the environment around a tumor, including surrounding blood vessels, immune cells, fibroblasts, signaling molecules, and extracellular matrix. Tumors and their microenvironment are closely related and constantly interact with each other with dynamic reciprocity. Tumor progression is influenced by the interactions between cancer cells and their environment, shaping therapeutic response and resistance.

[0028] The term "biomarker" as used herein refers to a metabolite or small molecule derived therefrom that is differentially present (i.e., increased or decreased) in a biological sample from a subject or group of subjects having a first phenotype (e.g., diseased) compared to a biological sample from a subject or group of subjects having a second phenotype (e.g., disease-free). A biomarker may be differentially present at any level, but generally will be at least 5%, at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 100%, at least 110%, at least 120%, at least 130%, at least 140%, at least 150%, at least 160%, at least 170%, at least 180%, at least 190%, at least 200%, at least 210%, at least 220%, at least 230%, at least 240%, at least 250%, at least 260%, at least 270%, at least 280%, at least 290%, at least 300%, at least 310%, at least 320%, at least 330%, at least 340%, at least 350%, at least 360%, at least 370%, at least 380%, at least 390%, at least 400%, at least 400%, at least 400%, at least 400%, at least 400%, at least 400%, at least 400%, at least 500%, at least 500%, at least 500%, at least 500%, at least The biomarkers are preferably present at levels that are at least 140%, at least 150%, or more increased, or generally at levels that are at least 5%, at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, or 100% (i.e., absent). Preferably, the biomarkers are differentially present at a statistically significant level.

[0029] The term "level" as used herein refers to the level of one or more biomarkers and means the absolute or relative amount or concentration of the biomarker in a sample.

[0030] The term "reference characteristic" as used herein refers to a metabolic characteristic indicative of a healthy subject or one or more of a medical condition, disease, or physical disorder. Within the reference characteristic, there is a reference level of one or more biomarkers (metabolites derived therefrom or small molecules derived therefrom), which may be the absolute or relative amount or concentration of one or more biomarkers, the presence or absence of one or more biomarkers, the range of amount or concentration of one or more biomarkers, the minimum and / or maximum amount or concentration of one or more biomarkers, the average amount or concentration of one or more biomarkers, and / or the median amount or concentration of one or more biomarkers.

[0031] As used herein, the term "statistically significant" refers to at least about a 95% confidence level, preferably at least about a 97% confidence level, more preferably at least about a 98% confidence level, and most preferably at least about a 99% confidence level, as determined using parametric or nonparametric statistics, such as, but not limited to, analysis of variance (ANOVA) or Wilcoxon rank sum test, expressed as p<0.05 for at least about a 95% confidence level in the Wilcoxon rank sum test. As used herein, the term "immune checkpoint blockade" refers to therapies that generally focus on the termination of immune responses by inhibiting immune suppressive molecules, thereby preventing the termination of immune responses, or activating T lymphocytes that are depleted during an immune response.

[0032] Whenever the term "at least," "greater than," or "greater than or equal to" precedes the first number in a series of two or more numbers, the term "at least," "greater than," or "greater than or equal to" applies to each and every number in the series. For example, 1, 2, or 3 or more is equivalent to 1 or more, 2 or more, or 3 or more.

[0033] Whenever the terms "no more than," "less than," or "less than or equal to" appear before the first number in a series of two or more numbers, the terms "no more than," "less than," or "less than or equal to" apply to each and every number in the series. For example, 3, 2, or 1 or less is equivalent to 3 or less, 2 or less, or 1 or less.

[0034] The use of the words "a" or "an" when used in conjunction with the term "comprising" in the claims and / or specification may mean "one," but it is also consistent with the meaning of "one or more," "at least one," and "one or more than one."

[0035] Although the term "or" in the claims is used to refer to alternatives only, or to mean "and / or" unless the alternatives are expressly indicated to be mutually exclusive, the present disclosure supports a definition that refers to alternatives only and "and / or." As used herein, "another" may mean at least a second or more.

[0036] The terms "comprise," "have," and "include" are open-ended linking verbs. Any form or tense of one or more of these verbs, such as "comprises," "comprising," "has," "having," "includes," "including," etc., also means open-ended. For example, any method that "comprises," "has," or "includes" one or more steps is not limited to possessing only those one or more steps, but also encompasses other unlisted steps.

[0037] III. Immunotherapy Treatment and Immune System Response Mechanisms A. The Tumor Microenvironment The immune system detects a wide variety of antigens, including viruses, parasites, allergens, and cancers, and initiates a response in the body against abnormal substances, cells, and / or tissues. Cancer growth, including malignant tumor growth, can also be recognized by a subject's immune cells to trigger an immune response. Immune cell activation can trigger numerous intracellular signaling pathways that require tight control to initiate an appropriate immune response. Cancer growth can closely interact with their microenvironment. Tumors can be composed of a heterogeneous population of cancer cells as well as a variety of resident and infiltrating host cells, secreted factors, and extracellular matrix proteins. Cancer and tumor progression can be significantly influenced by the interactions of cancer cells with this tumor microenvironment, which may ultimately determine tumor eradication, metastasis, therapeutic response, or therapeutic resistance. Mechanisms of the tumor microenvironment during cancer progression can provide therapeutic avenues in targeting components of the tumor microenvironment, such as immune checkpoint inhibitor therapy.

[0038] The tumor microenvironment, especially in solid tumors, may remain unhostile to immune cells such as effector T cells. Continuous concentration of immunosuppressive signals and lack of essential nutrients in the tumor microenvironment may lead to exhaustion of T cells. Overcoming the tumor microenvironment and determining early predictive response to treatment may be a key factor in improving the efficiency of immunotherapy in eradicating cancer cells in tumors. Metabolic reprogramming and adaptability to adapt to the rapid proliferation of cancer cells may be an important mechanism of therapeutic resistance in malignant cancer. Several immune cell types, including but not limited to macrophages, B cells, T cells, neutrophils, and dendritic cells, can reside in the tumor microenvironment and retain an active role in cancer progression.

[0039] B. Tumor Escape Mechanisms The progression of neoplasia to malignancy may occur in part due to failure of immune surveillance. Cancer cells may escape immune recognition and elimination, creating an immunosuppressive microenvironment. Due to the high consumption of cancer cells, natural immune cells in this area may face a nutrient-poor environment. Multiple metabolic by-products of cancer cell metabolism, such as lactate and glycolysis end products, may be toxic to natural immune cells, impairing their differentiation, activation, health, antitumor function, and rendering them unable to compete extensively with cancer cells.

[0040] Metabolic changes within the tumor microenvironment, such as hypoxia, can also affect the differentiation program of myeloid cells and alter their antigen-presenting properties. Hypoxia-mediated expression can selectively upregulate the expression of inhibitory ligands that promote immune suppression of T cells. Because cancer-mediated metabolic changes within the tumor microenvironment affect the cellular composition and function of the immune microenvironment, targeting the metabolic changes of cancer cells can affect the growth and progression of cancer cells, as well as provide a therapeutic target for improving antitumor immunity by altering the metabolic program of immune cells and their antitumor functions.

[0041] C. Immunotherapy Metabolic processes can modulate immune cell responses in quiescent conditions as well as during pathogenic processes such as infection, inflammation, cancer, and autoimmunity. In these complex conditions, immunotherapy may provide new therapeutic avenues. Macrophages as well as other immune cells exhibit metabolic flexibility depending on the disease pathology. Tumor-infiltrating lymphocytes are a notable part of the tumor microenvironment and may be associated with improved prognosis and response to therapy (Cogdill, Andrews, and Wargo 2017 Tomioka et al. 2018).

[0042] Immunotherapy can activate a subject's immune system to fight cancer. To efficiently eradicate cancer cells by immunotherapy, T cells or other immune cells can recognize tumor peptides presented by human leukocyte antigens (HLA). HLA or major histocompatibility complexes can be proteins involved in antigen presentation and can be encoded by HLA genes. Checkpoint blockade therapy has demonstrated meaningful antitumor activity with the subject's response influenced by various biological factors, including the complex interplay between the tumor, the tumor microenvironment, and the immune system (Hodi et al. 2010;Larkin, Ho and Wolchok 2015 Hugo et al. 2016;Ribas et al. 2016;Wolchok et al. 2017).

[0043] Immune checkpoint blockade therapy may be utilized to promote or inhibit T cell activation. An immune response may include an initiation phase and an activation phase in which the immune system recognizes a danger signal and is activated by innate signals to fight the danger. This response may be one of the first phases to resist infection or cancer, but it needs to be cut off as soon as the danger is controlled, as the continuation of this activation may cause tissue damage. The activation of the immune system is followed by a termination phase, in which endogenous immune suppressive molecules shut down the immune response to contain the damage. In cancer immunotherapy, therapeutic approaches have classically enhanced the initiation and activation of immune responses to increase the emergence and potency of T lymphocytes against cancer. Immune checkpoint blockade therapy may focus on the termination of the immune response by inhibiting immune suppressive molecules, thereby preventing the termination of the immune response or arousing T lymphocytes exhausted during the immune response. Blocking negatively regulatory immune checkpoints can restore exhausted immune cells' ability to kill infiltrating cancer and drive surviving cancer cells into a dormant state.

[0044] Immune checkpoints may be costimulatory and inhibitory elements inherent to the immune system. Immune checkpoints may help maintain self-tolerance and regulate the duration and tolerance of physiological immune responses to prevent damage to tissues when the immune system responds to pathogenic infections. Immune responses can also be initiated when T cells recognize antigens characteristic of tumor cells. The balance between costimulatory and inhibitory signals may be used to control immune responses from T cells, which may be regulated by immune checkpoint proteins. After T cells mature and become activated in the thymus, they can migrate to sites of inflammation and injury to perform repair functions. T cell function can occur either by direct action or through the recruitment of cytokines and membrane ligands involved in the immune system. The steps involved in T cell maturation, activation, proliferation, and function can be regulated through costimulatory and inhibitory signals, i.e., immune checkpoint proteins. Tumors may dysregulate the function of checkpoint proteins as an immune resistance mechanism. Thus, the development of modulators of checkpoint proteins may be of therapeutic value. Non-limiting examples of immune checkpoint molecules include CTLA4 and PD-1. These checkpoint molecules can function upstream of IL-2 in the pathway. IV. Examples of Biomarkers Used to Predict Immune System Response to Immunotherapy

[0045] Immunological checkpoint molecules may be members of the immunoglobulin superfamily or inhibitory receptors that suppress uncontrolled immune responses. Adaptive immune responses may be controlled by such checkpoint molecules, which can be used to maintain self-tolerance and minimize collateral tissue damage that may occur during immune responses. Numerous biomarkers of response to immune checkpoint blockade have been proposed, including PD-L I expression, interferon (IFN)γ-based signatures, tumor mutation burden, microsatellite instability (MSI) and mismatch repair deficiencies, genetic alterations including mutations within the antigen presentation machinery, loss of HLA heterozygosity, and diversity of T-cell repertoire (Herbst et al. 2014;Gao et al. 2016;Zaretsky et al. 2016;Roh et al. 2017 Sade-Feldman et al. 2017;Mariathasan et al. 2018;Chowell et al. 2019).

[0046] The diversity of biological traits that can affect response to immune checkpoint blockade therapy has led to expanding efforts to integrate multiple biological traits to identify biomarkers that better predict response to immunotherapy (Charoentong et al. 2017). In one such effort, a combined signature of purity-corrected tumor mutation burden, along with receptor tyrosine kinase (RTK) mutations, HLA mutations, and smoking signatures, was used to predict immune checkpoint blockade response in non-small cell lung cancer (NSCLC) (Anagnostou et al. 2020), and a melanoma study combined genomic, transcriptomic, and clinical data to predict response to immune checkpoint blockade (Liu et al. 2019).

[0047] Neoantigens can constitute a subset of somatic mutations that can be loaded onto the surface of MHC class I and class II molecules and presented to T cells. These neoantigens can be recognized by the immune system as endogenous tumor-specific (non-self) targets. Immune checkpoint blockade is thought to harness the ability of cytotoxic (CD8+) T cells to detect and destroy cancer cells that display neoantigens on the surface of h-IC class I molecules (Schumacher and Schreiber 2015). An integrated study of immunogenicity and neoantigen clonal structure predicted response and prognosis to immune checkpoint blockade in subjects with melanoma, lung cancer, and kidney cancer, suggesting broad applicability of the biomarker (Lu et al. 2020).

[0048] Recently, there has been an increasing effort to identify surrogate biomarkers for cancer diagnosis and progression using gene expression analysis, metabolomic, and proteomic approaches. Gene expression analysis can provide insight into crossover chromosomal events that can result in loss of heterozygosity, i.e., loss of entire genes and surrounding chromosomal regions. Loss of heterozygosity may indicate the absence of functional tumor suppressor genes in the lost region within the cancer. The tumor suppressor genes may be inactivated by this loss or by point mutations that leave no tumor suppressor genes to protect the body from cancer growth. HLA loss of heterozygosity detection may be a versatile cancer biomarker. V. Techniques for Generating Composite Biomarker Scores

[0049] As described herein, the composite biomarker score generated by the immunogenomics analysis system can incorporate information on adverse events (e.g., loss of HLA heterozygosity) within the antigen presentation machinery with predicted neoantigens to stratify subject response to immunotherapy. The composite biomarker score outperforms traditional single analyte biomarkers, suggesting that complex models capturing multiple aspects of tumor evasion can provide more robust stratification of subject response. Moreover, such data-intensive biomarkers are clinically actionable with comprehensive tumor characterization in various clinical populations achieved with limited tumor tissue. These findings provide not only accurate composite biomarkers of response for late-stage cancer patients, but also evidence supporting the use of whole-exome and transcriptome data in clinical settings. A. Generation of Genomic and Transcriptomic Data 1. Biological Samples

[0050] 1 illustrates an example of a schematic 100 for generating genomic and transcriptomic data from a biological sample, according to some embodiments. For example, the schematic 100 includes selecting a biological sample from a subject that includes cancer cells. In some cases, a pre-treatment normal blood sample and a tumor sample are taken from the subject. For example, the pre-treatment normal blood sample and the tumor sample may be taken from a subject with unresectable stage III / IV melanoma who has received anti-PD-1 therapy.

[0051] The biological sample can be processed to generate a subject's immunogenomic signature, which may include comprehensive tumor mutation information, gene expression quantification, neoantigen characterization, HLA (typing, mutations, and loss of heterozygosity), T-cell receptor repertoire characterization, microsatellite instability detection, cancer virus identification, and tumor microenvironment characterization. The signature data may then be analyzed along with clinical outcomes and a composite biomarker score calculated for the subject to identify a predictive level of responsiveness to a particular immunotherapy treatment.

[0052] Samples may be taken from a subject. Samples may be obtained from or include blood (e.g., whole blood), plasma, serum, umbilical cord blood, chorionic villi, amniotic fluid, lavage fluid (e.g., from bronchoalveolar, stomach, peritoneum, duct, ear, arthroscope), biopsy sample (e.g., from pre-implantation embryo), abdominal paracentesis sample, fetal nucleated cells or fetal cell debris, bile, breast milk, urine, saliva, mucosal discharge, sputum, stool, sweat, vaginal secretions, body fluid pool (e.g., testis), vaginal washings, pleural fluid, peritoneal fluid, cerebrospinal fluid, bronchoalveolar lavage fluid, nipple secretions, aspirates from various parts of the body (e.g., thyroid, breast), tears, embryonic cells, or fetal cells (e.g., placental cells). In some embodiments, blood samples are obtained by heel or finger prick, scalp vein, or ear lobe puncture. The biological sample may be a body fluid sample or a tissue sample (e.g., a skin sample). The biological sample may include any tissue or material from a living or dead subject. The biological sample may be an acellular sample. The biological sample may include protein or nucleic acid (e.g., DNA or RNA or fragments thereof). The sample may be fixed or unfixed. The sample may be embedded or free. The sample may be a formalin-fixed paraffin embedded sample.

[0053] The biological sample may include one or more nucleic acid molecules. The nucleic acid molecule may be a DNA molecule, an RNA molecule (e.g., mRNA, cRNA, or miRNA), and a DNA / RNA hybrid molecule. Examples of DNA molecules include, but are not limited to, double-stranded DNA, single-stranded DNA, single-stranded DNA hairpin, cDNA, and genomic DNA. The nucleic acid may be an RNA molecule, such as double-stranded RNA, single-stranded RNA, ncRNA, RNA hairpin, and mRNA. Examples of ncRNA include, but are not limited to, siRNA, miRNA, snoRNA, piRNA, tiRNA, PASR, TASR, aTASR, TSSa-RNA, snRNA, RE-RNA, uaRNA, x-ncRNA, hY RNA, usRNA, snaR, and vtRNA.

[0054] 2. Sequencing Whole-exome library preparation and sequencing may be performed to generate DNA sequences corresponding to genomic data from biological samples. DNA is extracted from the biological samples, processed, and subjected to whole-exome sequencing. Whole-exome capture libraries may be constructed using DNA from tumor samples and normal blood samples. In some cases, targeted probes are used to increase the inclusion rate of biomedically and clinically relevant genes. The procedure may be modified to obtain an average library insert length of approximately 250 bp. Sequencing reads undergo quality control processing (e.g., by FastQC) to provide FASTQ files. The FASTQ files are aligned to a reference genome to generate BAM files.

[0055] Transcriptome sequencing may be performed to generate RNA sequences corresponding to transcriptome data from biological samples. In some cases, transcriptome sequencing includes microarrays and RNA-Seq. Microarrays can be configured to measure the abundance of a defined set of transcripts through hybridization to complementary probe arrays. RNA-Seq may refer to the sequencing of the complementary DNA of transcripts in biological samples, where the abundance of complementary DNA is derived from the count value of each transcript.

[0056] In some cases, sample processing includes nucleic acid sample processing and subsequent nucleic acid sample sequencing. Some or all of the nucleic acid sample can be sequenced to provide sequence information, which may be stored or otherwise maintained in an electronic, magnetic, or optical repository. The sequence information can be analyzed with the aid of a computer processor, and the analyzed sequence information may be stored in an electronic repository. The electronic repository may contain a repository or collection of sequence information generated from the nucleic acid sample and the analyzed sequence information.

[0057] In some embodiments, whole genome sequencing may be used. In some cases, whole genome sequencing is used to identify variants in humans. In some cases, sequencing may include deep sequencing of a portion of the genome. For example, the portion of the genome may be at least about 50; 75; 100; 125; 150; 175; 200; 225; 250; 275; 300; 350; 400; 450; 500; 550; 600; 650; 700; 750; 800; 850; 900; 950; 1,000; 1,100; 1,200; 1,300; 1,400; 1,500; 1,600; 1,700; ,800;1,900;2,000;3,000;4,000;5,000;6,000;7,000;8,000;9,000;10,000;15,000;20,000;30,000;40,000;50,000;60,000;70,000;80,000;90,000;100,000, or more bases or base pairs. In some cases, the genome may be sequenced over 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more than 10 million bases or base pairs. In some cases, the genome may be sequenced over the entire exome (e.g., whole exome sequencing). In some cases, deep sequencing may include obtaining multiple reads across a portion of a genome. For example, obtaining multiple reads may include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1000, 10,000 reads, or more than 10,000 reads across a portion of a genome.

[0058] In some embodiments, the method may include detecting low allele fractions by deep sequencing. In some cases, deep sequencing is performed by next generation sequencing. In some cases, deep sequencing is performed by avoiding error-prone regions. In some cases, error-prone regions may include regions near sequence duplications, regions of abnormally high or low %GC, regions near homopolymers, regions near di- and trinucleotides, and other short repeats. In some cases, error-prone regions may include regions that cause DNA sequencing errors (such as polymerase slippage in homopolymer sequences).

[0059] In some embodiments, the method may include performing one or more sequencing reactions on one or more nucleic acid molecules in the sample. In some embodiments, one or more, two or more, three or more, four or more, five or more, six or more, seven or more, eight or more, nine or more, ten or more, fifteen or more, twenty or more, thirty or more, forty or more, fifty or more, sixty or more, seventy or more, eighty or more, ninety or more, one hundred or more, two hundred or more, three hundred or more, four hundred or more, five hundred or more, six hundred or more, seven hundred or more, eight hundred or more, nine hundred or more, or one thousand or more sequencing reactions are performed on one or more nucleic acid molecules in the sample. The sequencing reactions may be performed simultaneously, sequentially, or combinations thereof. The sequencing reactions may include whole genome sequencing or exome sequencing. The sequencing reactions may include a Maxam-Gilbert chain termination system or a high throughput system. Alternatively or additionally, the sequencing reaction may include Helioscope™ single molecule sequencing, Nanopore DNA sequencing, Lynx Therapeutics' massively parallel code sequencing (MPSS), 454 Pyrosequencing, single molecule real-time (RNAP) sequencing, Illumina (Solexa) sequencing, SOLiD sequencing, Ion Torrent™, ion semiconductor sequencing, single molecule SMRT(™) sequencing, polony sequencing, DNA nanoball sequencing, VisiGen Biotechnologies' techniques, or combinations thereof. Alternatively or additionally, the sequencing reaction may include one or more sequencing platforms including, but not limited to, Single Molecule Real Time (SMRT™) technologies such as the Genome Analyzer IIx, HiSeq, and MiSeq from Illumina, the PacBio RS system from Pacific Biosciences (California), and True Single Molecule Sequencing (tSMS™) technologies such as the Solexa Sequencer, and the HeliScope™ Sequencer from Helicos Inc. (Cambridge, MA).The sequencing reaction may also include an electron microscope or a chemically sensitive field effect transistor (chemFET) array. In some aspects, the sequencing reaction includes capillary sequencing, next generation sequencing, Sanger sequencing, sequencing by synthesis, sequencing by ligation, sequencing by hybridization, single molecule sequencing, or a combination thereof. Sequencing by synthesis may include reversible transcription termination sequencing, continuous single molecule sequencing, sequential flow sequencing, or a combination thereof. Sequential flow sequencing may include pyrosequencing, pH-mediated sequencing, semiconductor sequencing, or a combination thereof.

[0060] Some embodiments may include performing at least one long-read sequencing reaction and at least one short-read sequencing reaction. The long-read sequencing reaction and / or the short-read sequencing reaction may be performed on at least a portion of the subset of nucleic acid molecules. The long-read sequencing reaction and / or the short-read sequencing reaction may be performed on at least a portion of two or more subsets of nucleic acid molecules. Both the long-read sequencing reaction and the short-read sequencing reaction may be performed on at least a portion of one or more subsets of nucleic acid molecules.

[0061] Sequencing one or more nucleic acid molecules or a subset thereof may include sequencing at least about 5;10;15;20;25;30;35;40;45;50;60;70;80;90;100;200;300;400;500;600;700;800;900;1,000;1,500;2,000;2,500;3,000;3,500;4,000;4,500;5,000;5,500;6,000;6,500;7,000;7500;8, 000;8,500;9,000;10,000;25,000;50,000;75,000;100,000;250,000;500,000;750,000;10,000,000;25,000,000;50,000,000;100,000,000;250,000,000;500,000,000;750,000,000;1,000,000,000 or more sequence reads.

[0062] The sequencing reaction may comprise determining whether at least about 50;60;70;80;90;100;110;120;130;140;150;160;170;180;190;200;210;220;230;240;250;260;270;280;290;300; This may include sequencing 325;350;375;400;425;450;475;500;600;700;800;900;1,000;1,500;2,000;2,500;3,000;3,500;4,000;4,500;5,000;5,500;6,000;6,500;7,000;7,500;8,000;8,500;9,000;10,000;20,000;30,000;40,000;50,000;60,000;70,000;80,000;90,000;100,000, or more bases or base pairs. The sequencing reaction may comprise determining whether at least about 50;60;70;80;90;100;110;120;130;140;150;160;170;180;190;200;210;220;230;240;250;260;270;280;290;300; The method may include sequencing 325; 350; 375; 400; 425; 450; 475; 500; 600; 700; 800; 900; 1,000; 1,500; 2,000; 2,500; 3,000; 3,500; 4,000; 4,500; 5,000; 5,500; 6,000; 6,500; 7,000; 7,500; 8,000; 8,500; 9,000; 10,000; 20,000; 30,000; 40,000; 50,000; 60,000; 70,000; 80,000; 90,000; 100,000, or more consecutive bases or base pairs.

[0063] Preferably, the sequencing techniques used in the methods of the invention generate at least 100 reads per run, at least 200 reads per run, at least 300 reads per run, at least 400 reads per run, at least 500 reads per run, at least 600 reads per run, at least 700 reads per run, at least 800 reads per run, at least 900 reads per run, at least 1,000 reads per run, at least 5,000 reads per run, at least 10,000 reads per run, at least 50,000 reads per run, at least 100,000 reads per run, at least 500,00 reads per run, or at least 1,000,000 reads per run. Alternatively, the sequencing techniques used in the methods of the invention generate at least 1,500,000 reads per run, at least 2,000,000 reads per run, at least 2,500,000 reads per run, at least 3,000,000 reads per run, at least 3,500,000 reads per run, at least 4,000,000 reads per run, at least 4,500,000 reads per run, or at least 5,000,000 reads per run.

[0064] Preferably, the sequencing techniques used in the methods of the invention may generate at least about 30 base pairs, at least about 40 base pairs, at least about 50 base pairs, at least about 60 base pairs, at least about 70 base pairs, at least about 80 base pairs, at least about 90 base pairs, at least about 100 base pairs, at least about 110 base pairs, at least about 120 base pairs, at least about 150 base pairs, at least about 200 base pairs, at least about 250 base pairs, at least about 300 base pairs, at least about 350 base pairs, at least about 400 base pairs, at least about 450 base pairs, at least about 500 base pairs, at least about 550 base pairs, at least about 600 base pairs, at least about 700 base pairs, at least about 800 base pairs, at least about 900 base pairs, or at least about 1,000 base pairs per read. Alternatively, the sequencing techniques used in the methods of the invention may generate long sequencing reads. In some cases, the sequencing technology used in the methods of the invention is at least about 1,200 base pairs per read, at least about 1,500 base pairs per read, at least about 1,800 base pairs per read, at least about 2,000 base pairs per read, at least about 2,500 base pairs per read, at least about 3,000 base pairs per read, at least about 3,500 base pairs per read, at least about 4,000 base pairs per read, at least about 4,500 base pairs per read, 5,000 base pairs per read, at least about 6,000 base pairs per read, base pairs, at least about 7,000 base pairs per read, at least about 8,000 base pairs per read, at least about 9,000 base pairs per read, at least about 10,000 base pairs per read, 20,000 base pairs per read, 30,000 base pairs per read, 40,000 base pairs per read, 50,000 base pairs per read, 60,000 base pairs per read, 70,000 base pairs per read, 80,000 base pairs per read, 90,000 base pairs per read, or 100,000 base pairs per read may be generated.

[0065] High throughput sequencing systems allow for detection of sequenced nucleotides immediately or at the time of their incorporation into the growing strand, i.e., detection of sequences in real time or substantially real time. In some cases, high throughput sequencing produces at least 1,000, at least 5,000, at least 10,000, at least 20,000, at least 30,000, at least 40,000, at least 50,000, at least 100,000, or at least 500,000 sequence reads per hour, each read being at least 50 bases, at least 60 bases, at least 70 bases, at least 80 bases, at least 90 bases, at least 100 bases, at least 120 bases, at least 150 bases, at least 200 bases, at least 250 bases, at least 300 bases, at least 350 bases, at least 400 bases, at least 450 bases, or at least 500 bases per read. Sequencing may be carried out using a nucleic acid as described herein, such as genomic DNA, cDNA derived from an RNA transcript, or RNA as a template. 3. Alignment

[0066] The sequence reads (e.g., DNA sequences, RNA sequences) generated by the above sequencing techniques can be mapped to the corresponding reference genome (e.g., hs37d5 reference genome construction). In some cases, an alignment pathway performs alignment, duplication removal, and base quality score recalibration to generate genome and transcriptome data. This pathway uses Picard toolkit (RRID: SCR_006525) for duplication removal, and Genome Analysis Toolkit (GATK, RRID: SCR_001876) to improve sequence alignment and correct base quality score (BQSR). The aligned sequence data is then converted back to BAM format according to SAM (RRID: SCR_01095) specifications. In some cases, somatic variants are identified based on the alignment of sequence reads to the reference genome.

[0067] In some cases, whole transcriptome sequencing is aligned using STAR (RRID: SCR_015899) to calculate normalized expression values ​​in transcripts per million (TPM). Quality control of RNA sequencing and alignment may identify the following metrics: average read length, percentage of uniquely mapped reads, average length of mapped read pairs, number of splice sites, mismatch rate per base, deletion / insertion rate per base, average length of deletions / insertions, and alignment of aberrant read pairs including interchromosomal and orphan reads. B. Transcriptomic metrics derived from transcriptomic data

[0068] The immunogenomic analysis system processes the transcriptomic data corresponding to the biological sample to generate a set of transcriptomic indices. Each of the set of transcriptomic indices may represent one or more features corresponding to a set of peptides translated from a corresponding RNA sequence of the one or more RNA sequences. In some cases, the set of transcriptomic indices includes: (i) a quantitative or classifying indices representative of a predicted neoantigen load of the biological sample; (ii) a quantitative or classifying indices representative of one or more features of each of the one or more candidate neoantigens detected in the biological sample; (iii) a quantitative or classifying indices representative of one or more features of each of the one or more HLA proteins for which loss of cell surface presentation is detected; (iv) a quantitative or classifying indices representative of one or more features corresponding to HLA genes encoding the one or more HLA proteins for which loss of cell surface presentation is detected; and (v) a quantitative or classifying indices representative of the expression levels of one or more T cell receptors detected in the biological sample. For HLA proteins for which loss of cell surface presentation is detected, the corresponding indices can be generated by fitting the genomic and transcriptomic data to a neoantigen presentation predictive machine learning model.

[0069] 1. Signs of immune infiltration The set of transcriptome indices may include a quantitative or classifier index that represents the expression level of sequences corresponding to immune cells. In some cases, the quantitative or classifier index is an immune infiltration score, which is derived based on the amount of various tumor-infiltrating immune cell types. The immune infiltration score can be calculated using transcriptome data. For example, a semi-quantitative score that represents the enrichment of a set of genes may be calculated in a single sample. In some cases, the immune infiltration score is generated using a set of reference gene expression signatures that represent 17 cell types, including malignant cells, CAFs, endothelial cells, NK cells, B cells, macrophages, and CD8 + T cells and CD4 + T cells may also be included.

[0070] To generate an immune infiltration score, gene set enrichment analysis can be used to calculate a high enrichment score if a particular cell type-specific gene is expressed to the highest degree in the sample of interest (i.e., if the cell type is enriched in the sample), and a low enrichment score otherwise. Enrichment scores for the same cell type (gene set) can be compared across samples to define the immune infiltration of interest. Additionally or alternatively, immune infiltration scores are generated using deconvolution techniques that can quantitatively estimate the relative fraction of a cell type of interest (e.g., cancer cells). The deconvolution algorithm considers the gene expression profile of a heterogeneous sample as a superposition of gene expression levels of different cells and estimates the unknown cell fractions, which are then utilized into a signature matrix that describes the cell type-specific expression profile.

[0071] 2. Expression levels of T cell receptors The set of transcriptome indices may include quantitative or classification indices that represent the expression level of one or more T cell receptors detected from biological samples. The expression level of one or more T cell receptors can identify the level and distribution of clonal lymphocytes detected in biological samples. The quality and quantity of lymphocytes from biological samples can be used to identify various factors that affect the health and disease of subjects. The expression level of one or more T cell receptors can be interpreted as having normal immune diversity, proliferation, or reconstitution, or alternatively as having inflammation, infection, vaccination, autoimmunity, or cancer. In some cases, some analytical parameters used to evaluate the quality and quantity of lymphocyte infiltration of biological samples may include indices of diversity, concentration, uniformity, clonality, and entropy.

[0072] In some cases, the expression level of one or more T cell receptors corresponds to the clonality of T cell receptor beta (TCR-β) sequences detected in the biological sample. The immunogenomic analysis system processes the transcriptome data to characterize TCR-β clones that provide an expansion of the TCR-β range (approximately 100-fold expansion across the reference transcriptome). Non-productive clones with frameshifts or premature stop codons in the CDR3 sequences can be excluded, as well as low-confidence clones with alignment scores below the V or J hit threshold. Clonality may then be calculated as 1-Pielou uniformity.

[0073] 3. Differential Gene Expression The set of transcriptome indices may include a quantitative index that represents the number of reads per identified gene in the transcriptome data. For example, the number per million sequence reads can be calculated by normalizing the number of reads per gene by the total number of reads identified in the biological sample. In some cases, a threshold is selected as to whether a particular gene should be part of the quantitative index. For example, only genes with reads per million greater than 0 in 25% or more of the samples of the population can be included in the analysis. In some cases, the remaining data is processed using rlog transformation to analyze differential gene expression. Genes with an adjusted p-value less than 0.05 and a minimum log2 fold change less than -0.5 or greater than 1 are considered to be differentially expressed. The biological significance of differentially expressed genes can be identified at the pathway level using various gene sets, including but not limited to, MSigDB (Molecular Signatures Database, RRID: SCR_016863) authenticated gene sets and KEGG (RRID: SCR_012773) gene sets.

[0074] 4.Prediction of neoantigen presentation The set of transcriptomic signatures may include quantitative or classifying signatures representative of one or more characteristics of each of one or more HLA proteins for which loss of cell surface presentation is detected. In particular, the transcriptomic signatures may correspond to patient-specific tumor alterations that may prevent presentation of neoantigens, such as HLA mutations, loss of HLA heterozygosity, and β2-microglobulin mutations.

[0075] A neoantigen presentation predictive index can be generated by identifying candidate neoantigens generated using tumor-specific genomic events (single nucleotide mutations, indels, and fusions) that are validated using transcriptome data. All candidate peptides can be scored using a neoantigen presentation predictive machine learning model for predicting MHC class I presentation, which is trained using a large immune peptide data set. The trained neoantigen presentation predictive machine learning model can use data corresponding to each of the candidate peptides to generate an output predicting whether the candidate peptide will be presented and expressed on the cell surface. Based on the output of the machine learning model, a subset of candidate peptides that exceed a confidence threshold can be used to calculate a neoantigen burden score. To calculate a composite biomarker score, the neoantigen burden score can be adjusted to account for subject-specific tumor alterations that may impair neoantigen presentation, such as alterations to the MHC complex and antigen presentation machinery, and loss of HLA heterozygosity.

[0076] C. Genomic metrics derived from genomic data The immunogenomic analysis system can process the genomic data to generate a set of genomic indices. Each of the set of genomic indices can represent one or more characteristics corresponding to the corresponding DNA sequence as one or more DNA sequences. In some cases, the set of genomic indices includes (i) a quantitative or classifying indices representing one or more characteristics for each of one or more somatic mutations in the one or more DNA sequences; (ii) a classifying indices indicating whether loss of heterozygosity has occurred within at least one human leukocyte antigen (HLA) gene of the biological sample; and (iii) a quantitative or classifying indices representing a predicted tumor mutation load. For HLA loss of heterozygosity, the corresponding classifying indices can be generated by fitting the genomic data to an HLA deletion discrimination machine learning model.

[0077] 1. Single base mutations and indels The set of genomic indices may include quantitative or classification indices that represent one or more characteristics for each of one or more somatic mutations in one or more DNA sequences. The one or more somatic mutations may include single nucleotide mutations, insertion / deletion polymorphisms, copy number changes, and fusions in one or more nucleic acid molecules of the DNA sequence. In some cases, a quality index may be generated for each mutation identified in the DNA sequence, including the number of mutations, the ratio of transitions to transversions, the concordance rate of mutation levels, and the like. For example, the genomic data may be processed using a quality score recalibration module, which may stratify single nucleotide mutations according to their likelihood of representing false positive calls. In some cases, the sequence alignment information of the genomic data may be processed so that miscalled mutations can be corrected. Additionally or alternatively, the somatic single nucleotide mutation and insertion / deletion calls may be combined and analyzed through a set of tested filters based on 1) alignment indices such as sequence coverage and read quality, 2) positional characteristics such as proximity to gap regions, and 3) likelihood of being present in normal tissues.

[0078] 2. Loss of heterozygosity for specific HLA alleles The set of genomic indicators may also include a classification indicator indicating whether loss of heterozygosity has occurred in at least one HLA gene of the biological sample. Since loss of heterozygosity of HLA may affect neoantigen presentation, HLA loss of heterozygosity can be detected using HLA deletion discrimination machine learning models. Loss of heterozygosity of HLA may be considered as an acquired resistance mechanism that promotes immune escape by reducing the tumor neoantigen presentation ability to the immune system. Since the process of HLA loss is governed by selection pressures within the tumor microenvironment, especially at the later stages of tumor evolution, it was hypothesized that loss of heterozygosity of allele-specific HLA in a population of late-stage melanoma subjects may contribute to the reduced therapeutic response despite the elevated apparent neoantigen load.

[0079] To generate the above genomic index, biological samples may be processed using the following procedure: 1) map all tumor and normal reads to the allele-specific HLA of interest; 2) align homologous alleles to find all patient-specific mismatch sites; 3) calculate normalized b allele frequency and allele-specific inclusion rate at each mismatch site. For each gene, allele-specific features were input into an HLA deletion discrimination machine learning model to predict normalized b allele frequency and loss of heterozygosity including allele-specific mismatch sites, tumor purity, and tumor ploidy.

[0080] 3. Mutation Load The set of genomic indices may include quantitative or classifier indices that represent predicted tumor mutational burden. Tumor mutational burden would refer to the total number of mutations (alterations) found in the DNA of cancer cells. Recognizing tumor mutational burden may help to plan the best treatment, and tumor mutational burden has been identified as a potential biomarker for immune checkpoint blockade response.

[0081] D. Generation of a Composite Biomarker Score The immunogenomics analysis system generates a composite biomarker score derived from a set of genomic indicators and a set of transcriptomic indicators, and determines the predicted level of the subject's responsiveness to a certain type of immunotherapy treatment based on the composite biomarker score.For example, a composite biomarker score can be generated using the transcriptomic indicator corresponding to the neoantigen load score, and the composite biomarker score can be adjusted based on the predicted tumor mutation load identified from the genomic data.The composite biomarker score can therefore take into account the barriers to neoantigen presentation and other established resistance markers.Incorporating antigen presentation into the composite biomarker score can enhance the predictive level associated with immune checkpoint blockade response.

[0082] Although the measurement of neoantigen load can predict which subjects will benefit from immunotherapy by increasing the measurement, a composite biomarker score can be derived based on genomic and transcriptomic indicators corresponding to additional resistance mechanisms resulting from genetic mutations in antigen presentation mechanisms at both germline and somatic levels. These additional resistance mechanisms can further regulate immune response by reducing neoantigen presentation capacity. Thus, a composite biomarker can use indicators corresponding to neoantigen load as biomarkers, but can also include genomic and transcriptomic indicators corresponding to RNA expression levels as well as additional data derived from subsequent processing steps and longitudinal processing.

[0083] In some cases, the composite biomarker score corresponds to a neoantigen burden score that is adjusted to account for subject-specific tumor alterations that may further impede neoantigen presentation, including HLA mutations, loss of HLA heterozygosity, and β2-microglobulin mutations. As a result, analysis of a subject using a composite biomarker score can improve prediction of treatment outcome when compared with neoantigen and tumor mutation burdens individually. A composite biomarker approach that models both biological mechanisms and impaired neoantigen presentation may serve as a more powerful biomarker for immune checkpoint blockade therapy than many current biomarkers built on simpler biological models of tumor immune responses. Unlike approaches based on tumor mutation burden, a composite biomarker score can be generated by modeling a broader range of mechanisms of neoantigen presentation.

[0084] Additionally or alternatively, a subset of somatic mutations associated with reduced response to immunotherapy (e.g., HLA class I and B2M mutations, loss of heterozygosity of HLA class I genes) is weighted to adjust the composite biomarker score. By accounting for these escape mechanisms, the composite biomarker score can capture a more complete representation of tumor antigen presentation to the immune system, enhancing the predictive strength of this biomarker. The above approach can provide more accurate results when applied to one or more specific types of cancer, such as a population of subjects with non-small cell lung cancer and squamous cell carcinoma of the head and neck, as loss of heterozygosity of HLA is identified as a widespread escape mechanism affecting the progression of these types of cancer. For example, tumor data from head and neck cancer, lung adenocarcinoma, pancreatic cancer, and prostate cancer revealed allele-specific loss of expression at frequencies greater than 45%. Loss of heterozygosity of HLA combined with the emergence of somatic mutations of class I HLA genes can be captured by the composite biomarker score to identify damaging events to the antigen presentation machinery.

[0085] Thus, a composite biomarker score can integrate a set of broad biological characteristics for multiple dimensions, i.e., exome and transcriptome, tumor and immune response and resistance, and can then be used to predict immune checkpoint blockade response, reflecting the biological mechanisms driving response and resistance to immunotherapy.

[0086] E. Treatment Options The composite biomarker score can serve as a powerful predictor of immune checkpoint blockade therapy response. As shown in the figure, the composite biomarker score was able to more significantly separate responders and non-responders to immune checkpoint blockade therapy compared to tumor mutation burden and other single analytes / genes and expression signatures assayed in the detection population. The value of the composite biomarker score for predicting responsiveness to specific immunotherapies was further demonstrated by validating these findings in a large independent validation population.

[0087] The composite biomarker score can further demonstrate that neoantigens can induce immune responses and promote clinical responses to immunotherapy. Only a weak association was observed between response and tumor mutation burden, but a stronger association was found between neoantigen load and subject response. It has been suggested that this finding is due to the confounding effect of the distribution of melanoma subtypes within the patient population in various clinical studies, which may negatively affect the predictive power of tumor mutation burden. However, such population-related issues did not appear to affect neoantigen load. Improved robustness of neoantigen load as a biomarker may be achieved by including additional data from subsequent processing steps as well as RNA expression levels, since this measurement has been shown to correlate with protein representation in the MHC-bound peptide repertoire.

[0088] In some cases, additional factors have been identified that affect subject response beyond neoantigen load. As an illustrative example, the non-responding outliers with the highest composite biomarker scores observed in the detection population also contain high-impact nonsense PD-1 mutations, which can be interpreted as likely to prevent response to anti-PD1 therapy. Non-responding subjects who are outliers in the validation population with high composite biomarker scores correspond to subjects suffering from metastatic desmoplastic melanoma, which is associated with high levels of mutational load and distinct clinicopathological and genetic characteristics compared to typical cutaneous melanoma. Thus, clinical response data can be used in conjunction with the composite biomarker score to identify the level of heterogeneity in subject response to immunotherapy. Furthermore, when combined with the composite biomarker score of clinical response data, subsets of malignancies that are vulnerable to certain combinations of treatments can be identified. Finally, when combined with the composite biomarker score of clinical response data, other mechanisms of treatment resistance or response that extend beyond neoantigen presentation can be identified.

[0089] The composite biomarker score can thus be used to determine a treatment to prevent, halt, reverse, or ameliorate the disease. The disease can be cancer. The composite biomarker score can indicate a predicted level of responsiveness of the subject. The composite biomarker score can thus be output as a report that identifies (i) a recommendation for a particular treatment; (ii) a recommendation to administer a particular treatment to the human subject; and / or (iii) a recommendation not to administer a particular treatment to the human subject based on the predicted level of responsiveness of the subject to the particular treatment. In some embodiments, the recommended treatment is administered to the human subject.

[0090] Non-limiting examples of cancer include acute lymphoblastic leukemia, acute myeloid leukemia, adrenal cortical carcinoma, AIDS-related cancer, AIDS-related lymphoma, anal cancer, appendix cancer, trocytoma, neuroblastoma, basal cell carcinoma, bile duct cancer, bladder cancer, bone cancer, brain tumors such as cerebellar astrocytoma, cerebral astrocytoma / malignant glioma, ependymoma, medulloblastoma, supratentorial primitive neuroectodermal tumor, visual pathway glioma, and hypothalamic glioma, breast cancer, bronchial adenoma, Burkitt's lymphoma, cancer of unknown primary, central nervous system lymphoma, cerebellar astrocytoma, cervical cancer, childhood cancer, chronic Lymphocytic leukemia, chronic myeloid leukemia, chronic myeloproliferative disorders, colon cancer, cutaneous T-cell lymphoma, desmoplastic small round cell tumor, endometrial cancer, ependymoma, esophageal cancer, Ewing's sarcoma, germ cell tumors, gallbladder cancer, stomach cancer, gastrointestinal carcinoid tumors, gastrointestinal stromal tumors, glioma, hairy cell leukemia, head and neck cancer, heart cancer, hepatocellular (liver) cancer, Hodgkin's lymphoma, hypopharyngeal cancer, intraocular melanoma, pancreatic islet cell carcinoma, Kaposi's sarcoma, kidney cancer, laryngeal cancer, lip and oral cavity cancer, liposarcoma, liver cancer, lung cancer including non-small cell and small cell lung cancer, lymphoma leukemia, macroglobulin malignant bone / osteosarcoma fibrous histiocytoma, medulloblastoma, melanoma, mesothelioma, metastatic squamous cell neck cancer including occult oral cancer, multiple endocrine neoplasia syndrome, myelodysplastic syndrome, myeloid leukemia, nasal and paranasal sinus cancer, nasopharyngeal cancer, neuroblastoma, non-Hodgkin's lymphoma, non-small cell lung cancer, oral cavity cancer, oropharyngeal cancer, osteosarcoma / malignant fibrous histiocytoma, ovarian cancer, ovarian epithelial cancer, ovarian germ cell tumor, pancreatic cancer, pancreatic islet cell cancer, paranasal sinus and nasal cancer, parathyroid cancer, penile cancer, pharyngeal cancer, pheochromocytoma, pineal astrocytoma, pine These include: oocyte germinoma, pituitary adenoma, pleuropulmonary blastoma, plasma cell neoplasms, primary central nervous system lymphoma, prostate cancer, rectal cancer, renal cell carcinoma, transitional cell carcinoma of the renal pelvis and ureter, retinoblastoma, rhabdomyosarcoma, salivary gland cancer, sarcoma, skin cancer, Merkel cell skin cancer, small intestine cancer, soft tissue sarcoma, squamous cell carcinoma, gastric cancer, T-cell lymphoma, pharyngeal cancer, thymoma, thymic carcinoma, thyroid cancer, (gestational) trophoblastic neoplasm, carcinoma of unknown primary site, urethral cancer, uterine sarcoma, vaginal cancer, vulvar cancer, Waldenström's macroglobulinemia, and Wilms' tumor.Examples of diseases or conditions for which integrative composite biomarkers can be used include hematological malignancies, solid malignancies, metastatic cancers, and benign tumors.

[0091] Multiple subjects suffering from cancer can benefit from the use of integrated composite biomarkers. The subjects can be humans, non-human primates such as chimpanzees, and other apes and monkeys; livestock animals such as cows, horses, sheep, goats, and pigs; domestic animals such as rabbits, dogs, and cats; laboratory animals including rodents such as rats, mice, and guinea pigs, and the like. The subjects can be of any age. For example, the subjects can be elderly, adults, adolescents, pre-adolescents, children, infants, and babies.

[0092] A patient's health or treatment options may be assessed by providing a bodily fluid or tissue sample from the subject; collecting genomic and proteomic signatures from the bodily fluid or tissue sample; and comparing the genomic and proteomic signatures to at least one reference signature to assess the subject's health. The reference signature may define at least one characteristic of one or more diseases, injuries, or disorders. The reference signature may be constructed from genomic or proteomic signatures collected from subjects with the same disease, from a healthy population, or both. The method may include repeatedly comparing and monitoring the genomic or proteomic signature to the reference signature over time. Aspects of the present disclosure may include statistically analyzing the difference between the tumor signature and the reference signature to identify at least one biomarker. Biomarkers or biomarker groups with a significance level of less than 95%, 97%, 98%, or 99% may be eliminated.

[0093] In some aspects, the disclosure can provide a method of adaptive immunotherapy for the treatment of cancer in a subject, comprising administering to the subject a first course of a first immunotherapeutic compound; and obtaining comprehensive tumor and immune-related molecular information related to additional novel and investigational biomarkers, such as neoantigen load, HLA genotype diversity, loss of heterozygosity of HLA, immune repertoire profile, immune cell deconvolution, cancer viruses, etc., wherein the second course of immunotherapy includes a second immunotherapeutic compound if the tumor and immune-related molecular profile indicates an insufficient response to the first immunotherapeutic compound; or includes a second course of the first immunotherapeutic compound if the tumor and immune-related molecular profile does not indicate an insufficient response to the first immunotherapeutic compound. The one or more biological samples obtained after administering the first dose of the first immunotherapeutic compound of the first course may be obtained on the same day that the first immunotherapeutic compound of the first course is administered in a subsequent dose.

[0094] The treatment, test, or analysis may be provided to the subject prior to clinical onset of the disease. The treatment, test, or analysis may be provided to the subject after clinical onset of the disease. The treatment, test, or analysis may be provided to the subject 1 day, 1 week, 6 months, 12 months, or 2 years after clinical onset of the disease. The treatment, test, or analysis may be provided to the subject 1 day or more, 1 week or more, 1 month or more, 6 months or more, 12 months or more, or 2 years or more after clinical onset of the disease. The treatment, test, or analysis may be provided to the subject less than 1 day, 1 week, 1 month, 6 months, 12 months, or 2 years after clinical onset of the disease. The treatment, test, or analysis may also include treating, testing, or analyzing humans in a clinical trial. VI. Experimental Results for Generating Composite Biomarker Scores

[0095] To demonstrate the efficacy of the composite biomarker score in predicting immune system response to immunotherapy, the following experiment was performed. Pre-treatment formalin-fixed paraffin-embedded tumor and normal blood samples were collected and characterized in pairs to generate comprehensive tumor mutation information, gene expression quantification, neoantigen characterization, HLA (typing, mutation, and loss of heterozygosity), TCR repertoire characteristics, and tumor microenvironment characteristics. These data were then partitioned with clinical outcomes, where a composite neoantigen score was calculated for each subject along with additional biomarkers such as tumor mutation burden.

[0096] A. Swarm 1. Detection population Regarding the study population, 51 subjects with treated unresectable stage III / IV melanoma were retrospectively enrolled in a randomized or open-label fashion. Subjects were treated with either nivolumab (480 mg intravenous (IV) every 4 weeks or 240 mg IV every 2 weeks), a combination of nivolumab and ipilimumab (1 mg / kg IV and 3 mg / kg IV, respectively, every 3 weeks), or pembrolizumab (200 mg IV every 3 weeks). Solid tumor and blood samples were collected within 3 months before the start of treatment. Computed tomography scanning measurements were performed 10–12 weeks after the start of treatment, with follow-up scanning measurements every 3 months thereafter. Responders were defined as complete response (CR) or partial response (PR). Non-responders were defined as stable disease (SD) or progressive disease (PD).

[0097] 2. Test population Replication of the prediction results was performed using publicly available NGS data taken from advanced melanoma subjects who received immune checkpoint blockade therapy. Whole exome and RNA sequencing data for this study were obtained from dbGaP (NCBI database of genotypes and phenotypes, RRID: SCR_002709) (study accession number: phs000452.v3.p1). Subjects with confounded response to treatment (2 subjects) and subjects with low purity tumors (7 subjects) were excluded from the analysis, leaving evaluable subjects (110 subjects) for validation. Clinical characteristics of the validation population are provided in the original study.

[0098] 3. Statistical analysis of clinical data Concerning the generation of clinical data, progression-free survival (PFS) and overall survival (OS) were estimated using the Kaplan-Meier method. Objective response rates were recorded as proportions with Clopper-Pearson exact CIs. Associations between groups and categorical variables were tested using Fisher's exact test and chi-square test. When variance between three or more groups was considered, the Kruskal-Wallis H test was used. Pairwise comparisons of values ​​were performed using the Wilcoxon-Mann-Whitney rank sum test (MWW). The listed P values ​​were adjusted using the Benjamini-Hochberg correction. RNA pathway analysis was performed using the Kolmogorov-Smirnov (KS) statistic. Correlations between continuous variables were determined using Kendall's tau coefficient. Prediction models were generated using logistic regression, and the AUROC was used to determine the ability to distinguish response from non-response according to published methods (28). All tests were two-sided and FDR values ​​of <0.1 for pathway analyses and P values ​​of <0.05 for all other tests were considered statistically significant. The results are shown in the following table.

[0099] B. Clinical Data of the Population 1. Clinical characteristics of samples from the detection population For the 51 unresectable melanoma subjects in the immune checkpoint blockade treated population, the median follow-up time for the population was 24 months after treatment, and 33 of the 51 subjects (50%, 95% Clopper-Pearson confidence interval: 50-78%) had an objective response at first assessment by Response Evaluation Criteria in Solid Tumors (RECIST) 1.1. Within the clinical population, tumors arose from the head and neck region (31%), trunk (31%), extremities (25%), acral region (6%), mucosal region (4%), and occult region (2%). In addition to these data, sex, age, and other subject-specific demographic information are presented. The following table provides a summary of various characteristics related to the subjects in the clinical population.

[0100] [Table 1]

[0101] There were no statistically significant differences in objective response rates between sites of disease origin, as shown in Table 1. An additional 11 subjects (22%) had received prior treatment with a checkpoint inhibitor, whereas 40 subjects (78%) had not received immune checkpoint blockade. Subjects received either pembrolizumab (29 subjects, 57%), nivolumab (15 subjects, 29%), or the combination of nivolumab and ipilimumab (7 subjects, 14%).

[0102] 2. Genomic data corresponding to samples from the detection population

[0103] Examination of mutations associated with responding and non-responding tumors revealed no significant single-gene predictors of response after multiple hypothesis correction (subject-level mutation data). The table below shows log2 and p-values ​​providing comparative data between responders and non-responders for each gene identified in the clinical population.

[0104] [Table 2-1] [Table 2-2] [Table 2-3] [Table 2-4]

[0105] 3. Immune pathway data corresponding to population subjects We then determined the genetically disrupted pathways that corresponded to the clinical data. The most frequently disrupted pathways included the RTK-RAS and WNT pathways (disrupted in 73% and 51% of the population, respectively). Mutations were detected throughout the RTK-RAS pathway. A large number of RTKs were mutated, including ROS1 and ERBB4, as well as RAS family genes including NRAS, BRAF, MAPK1 and 2.

[0106] Figures 2A-B show statistical data corresponding to oncogenic alterations in genomic and transcriptomic data corresponding to subjects in a clinical population. Figure 2A shows mutations in known oncogenic pathways in late stage melanoma subjects. In Figure 2A, the proportion of pathways affected is shown by the number of genes mutated in the pathway (51 samples included in this analysis). Figure 2B shows a visualization of mutations occurring in the RTK-RAS pathway. Tumor suppressor genes are shown in red and oncogenes in blue. Dots represent the absence of mutations in a particular gene. Each column represents a tumor and green blocks represent mutations in a particular gene.

[0107] C. Transcriptome metrics 1. Differential Gene Expression Transcriptomic data was generated for each subject in the detection population. From the transcriptomic data, various transcriptomic indices were generated. For example, 121 differentially expressed genes were identified in the responding subjects (48 evaluable subjects; adjusted P value ≤ 0.05, log fold change > 2 or < -0.5). Figure 3A-C shows statistical data corresponding to the transcriptomic indices identifying differentially expressed genes associated with immune system response. Figure 3A shows the 50 genes with the highest levels of differential expression in this population, where fold changes are provided to compare responding and non-responding subjects. Figure 3A further shows Benjamini-Hochberg corrected P values ​​< 0.05 for each gene in the corresponding set of 48 genes. Although not all are shown in Figure 3A, enrichment was observed for 29 of these genes and reduced expression was observed for 92 genes. For illustration, Figure 3B shows a heat map of differentially expressed genes for each subject in the clinical population. In FIG. 3B, each column represents a subject and each row represents a gene.

[0108] Among the most strongly upregulated genes (log2 fold change = 3.28; FDR adjusted P value = 0.0005) was delta-like ligand 3 (DLL3), an inhibitory Notch ligand highly expressed in small cell lung cancer tissues and other tumor tissues. Due to its low cytoplasmic expression in normal tissues compared to its uniformly elevated cell surface expression in tumors, the delta-like ligand 3 gene is currently being explored as a potential therapeutic target. In addition, four members of the keratin (KRT) family (KRT72, 73, 81, 86), a group of genes identified with widespread associations with cancer development, showed altered expression levels when comparing responders and non-responders. Validation of gene expression analysis results for DLL3 and KRT family genes confirmed significance for DLL3 (MWW; P = 0.02), but not for KRT72, 73, 81, and 86 (MWW; P = 0.44, P = 0.41, P = 0.6, and P = 0.17, respectively). These differences in validation results may be due to reduced sensitivity in determining differential expression of individual genes.

[0109] Although there was no significant enrichment at the population level, IDO1 expression was detected at very high levels in three subjects (median IDO1 TPM = 10.36; outliers IDO1 TPM = 1955, 661, and 451). For illustration, Figure 3C shows a set of box plots comparing IDO1 gene expression levels in responding and non-responding subjects. Gene expression values ​​are shown in units of transcripts per million kilobases (TPM). In the group of responding subjects, outliers were identified in three subjects. Although IDO1 expression values ​​did not appear to be related to response to immunotherapy, these outliers with elevated expression levels may represent an escape mechanism that prevents a complete response within the corresponding subjects (48 subjects). For example, two of the subjects overexpressing IDO1 may not have achieved a complete response to immunotherapy, possibly due to an IDO1-driven immunosuppressive environment.

[0110] 2. Gene Enrichment Analysis Gene set enrichment analysis was then performed to identify pathways that were differentially regulated within the clinical population. Figure 4 shows the statistical data corresponding to the normalized enrichment scores for each differentially regulated immune pathway, where normalized enrichment scores were generated based on gene set enrichment analysis. As shown in Figure 4, significant enrichment of pathways related to immune function was identified in responder subjects with upregulated genes. Benjamini-Hochberg corrected P values ​​of less than 0.05 are shown. Inflammatory signaling cascades were present among the most highly enriched pathways identified (significance set as FDR less than 0.1). Activation of immune pathways may be attributed to other enriched pathways. For example, Th17 cell differentiation may be driven by (i) the cytokine TGF-β, which induces RORγt in Th17 cells; and (ii) IL-6, which induces the Th17 lineage. The observed Th17 enrichment may be positively regulated by the observed increase in STAT3 signaling, which helps promote Th17 differentiation.

[0111] 3. Expression levels of T cell receptors Figures 5A-C show statistical data corresponding to the transcriptome index identifying the expression levels of T cell receptors. The adaptive immune system can respond to a wide range of antigens due to its broad repertoire of unique T cell receptors (TCRs). The box plots in Figures 5A, B and C encompass the interquartile range from the 25th percentile at its lower limit to the 75th percentile at its upper limit, with the median indicated by a horizontal line. The upper whiskers include the maximum value within the 1.5-fold interquartile range above the 75th percentile. The lower whiskers include the minimum value within the 1.5-fold interquartile range below the 25th percentile. To characterize the pre-treatment tumor immune landscape, the diversity of the TCR-β repertoire was identified in a subset of subjects (28 subjects) in the population. Clonality was determined for the clonal abundance of all proliferating TCR-β sequences using 1-Pielou uniformity. Since intratumor heterogeneity is believed to be a determinant of immune response, the mutant allele tumor heterogeneity (MATH) score, which indicates the estimated level of tumor heterogeneity, was compared with the identified TCR-β clonality. Figure 5A shows a set of box plots identifying the comparison of TCR-β clonality between low and high levels of mutant allele tumor heterogeneity. As shown in Figure 5A, a significant association (MWW; P=0.014) was identified between high tumor heterogeneity and clonal diversity of the TCR-β repertoire.

[0112] FIG. 5B shows a set of box plots identifying a comparison of TCR-β clonality between a first group of subjects identified as responsive to immunotherapy and a second group of subjects identified as non-responsive to immunotherapy. As shown in FIG. 5B, TCR-β clonality is elevated in responding subjects compared to non-responding subjects (28 subjects; MWW; P=0.047). Therefore, TCR-β clonality can be considered to be significantly associated with treatment outcome. Furthermore, FIG. 5C shows a line graph identifying a comparison of progression-free survival probability between a first group identified as having high TCR-β clonality and a second group identified as having low TCR-β clonality. FIG. 5C shows that a significantly longer progression-free survival was observed in subjects with high clonality compared to subjects with low clonality (two-tailed KM log-rank test, P=0.0043), where high / low clonality stratification was calculated separately for old / young populations (using the median population age as the cutoff point).

[0113] 4.Immune infiltration sign Characterization of immune and stromal cell populations within the tumor microenvironment in a population was performed. The data obtained was used to generate a semi-quantitative immune infiltration score. Figure 6 shows a set of box plots identifying the comparison of enrichment scores between the first group of responders and the second group of non-responders. The comparison of enrichment scores was identified for various forms of tumor-infiltrating lymphocytes, including regulatory T cells (TREGs), natural killer cells (NK cells), and cancer-associated fibroblasts (CAFs). As shown in Figure 6, responders and non-responders shared a broadly similar distribution of immune cell expression across various forms. Thus, immune cell gene expression levels alone do not appear to be a strong predictor of the level of responsiveness to immunotherapy. However, as described herein, these expression levels can be a contributing factor in generating a composite biomarker score that accurately predicts responsiveness to immunotherapy.

[0114] 5. Neoantigen loading A strong correlation with responsiveness to immune checkpoint blockade can be achieved with a neoantigen-based biomarker approach. For this particular test example, two different neoantigen models were generated to compare their respective performance levels. The first neoantigen model corresponds to a score based on neoantigen load alone, and the second neoantigen model corresponds to an extension of the first model to take into account impairments to neoantigen presentation and other established resistance markers. The second neoantigen model thus corresponds to a model for generating a composite biomarker score.

[0115] We used features derived from exome and transcriptome data to calculate a neoantigen burden score. Putative neoepitopes were predicted from single-nucleotide mutations, indels, and fusions detected from both exome and transcriptome sequencing. To improve MHC class I neoantigen prediction, we generated mass spectrometry-based peptide binding data from monoallelic HLA transgenic cell lines. This data was used to train an improved machine learning algorithm that integrates HLA binding, proteasomal cleavage, and gene expression information to improve neoantigen prediction.

[0116] Figures 7A-B show statistical data corresponding to the transcriptomic index identifying neoantigen burden across different genes and disease sites. Figure 7A shows a set of box plots identifying neoantigen burden scores corresponding to driver mutations corresponding to the BRAF, NRAS, NF1, and WT genes. Figure 7A shows that neoantigen burden varies significantly between tumors with different driver mutations, with significant variation emerging between subtypes (Kruskal-Wallis, P=1e-04).

[0117] Additionally, Figure 7B shows a set of box plots identifying the neoantigen burden scores corresponding to the various disease sites of melanoma including extremities, head and neck, mucosa, trunk, and occult regions. In Figure 7B, no significant association across disease sites of origin was detected (Kruskal-Wallis, P=0.08). Thus, there was no overall variation in neoantigen burden when comparing tumors arising from different sites of origin, although significant variation was evident in post-hoc comparisons between melanomas of extremities and trunk regions (MWW; P=0.047).

[0118] 8A-F show statistical data identifying neoantigen burden scores across different subjects, where the neoantigen burden scores can predict the responsiveness of subjects treated with immunotherapy. FIG. 8A shows a set of box plots corresponding to a comparison of neoantigen burden scores between a first group of subjects who responded to immunotherapy and a second group of subjects who did not respond to immunotherapy. In FIG. 8A, each box plot encompasses an interquartile range from its lower limit of the 25th percentile to its upper limit of the 75th percentile, with the median indicated by a horizontal line. The upper whiskers include the maximum value within 1.5 times the interquartile range above the 75th percentile. The lower whiskers include the minimum value within 1.5 times the interquartile range below the 25th percentile. Neoantigen burden was found to be significantly higher in responders compared to non-responders (51 subjects; MWW; P=0.016). Figure 8B shows a set of box plots corresponding to a comparison of neoantigen burden scores of groups of subjects (e.g., responders, non-responders) within the validation population. The data from the validation population in Figure 8B confirms that subjects who responded to treatment showed significantly higher neoantigen burden (MWW; P=0.021).

[0119] Other forms of experimental data also show that high neoantigen load scores are more strongly associated with responsiveness to immunotherapy. Figure 8C shows a line graph identifying a comparison of progression-free survival probability between a first group identified as having high neoantigen load and a second group identified as having low neoantigen load. As shown in Figure 8C, a significantly longer progression-free survival period was observed in subjects with high neoantigen load compared to subjects with low neoantigen load (two-tailed KM log-rank test; P=0.002). Figure 8D shows a line graph identifying a comparison of progression-free survival probability between groups of subjects in a validation population, and Figure 8E shows a line graph identifying a comparison of overall survival rate between groups of subjects in a validation population. Figure 8D shows that within the validation population, progression-free survival in subjects with high neoantigen load was not significantly longer than in subjects with low neoantigen load (two-sided KM log-rank test, P = 0.085), whereas Figure 8E shows that a significant improvement in overall survival was observed in subjects with high neoantigen load (two-sided KM log-rank test, P = 0.005).

[0120] Figure 8F shows the receiver operating characteristic curve for identifying the performance level of the neoantigen burden score model. As shown in Figure 8F, the area under the curve value for the neoantigen burden score model was 0.71, and the cross-validation area under the curve value (mean) was 0.69 (log likelihood ratio P = 0.0329).

[0121] D. Genomic Indicators 1. Mutation characteristics In addition to the transcriptomic data, genomic data was generated for each subject in the detection population. From the genomic data, various genomic indices were generated. Figures 9A-F show statistical data identifying one or more characteristics associated with mutations present in each subject sample of the detection population. Figure 9A shows the identification of mutations in various genes in subjects receiving anti-PD-1 therapy. The box plots on top of Figure 9A represent the mutation burden. The tile plots show the mutated genes (rows) per sample (columns), with the color of the tile indicating the type of mutation. The box plots on the right represent the number of subjects with mutations in a particular gene, colored to indicate the type of mutation. In this tile plot, the first row represents the treatment response, either a response (partial or complete response; dark green; 33 subjects) or a non-response (black; 18 subjects).

[0122] In Figure 9A, the median nonsynonymous tumor mutational burden was 4.07 mutations / MB (interquartile range, 0.95-12.455). This genomic index appears to be consistent with values ​​observed in known datasets. For example, Figure 9B shows the mutation load identified in each sample across the various datasets. The mutation load levels in the detection population correspond to those in the TCGA-SKCM dataset (melanoma). In Figure 9B, each dot represents one sample and the red horizontal line is the median mutation load for each cancer type. The vertical axis (on a logarithmic scale) shows the number of mutations per sample.

[0123] Figure 9C is a set of box plots that identify the mutational load of each type of single nucleotide variant for each subject in the detection population, and a bar graph that shows the distribution of single nucleotide variant types. In Figure 9C, single nucleotide variants are classified as either transitions or transversions (n=49). The box plot on the left shows the overall distribution of six different substitution types, and the box plot on the right shows the distribution of transitions (Ti) and transversions (Tv). As shown in Figure 9C, C>T transitions appear to form the majority of the identified single nucleotide variants (76%).

[0124] Figure 9D shows a bar graph identifying the distribution of the three mutation signatures for each subject in the detection population. Signatures were extracted by decomposing the nucleotide substitution substrate and classified into 96 substitution classes based on the bases immediately surrounding the mutated base, and the population was classified by three major signatures. According to Figure 9D, the most commonly identified driver mutation was in BRAF, occurring in 33% of the subjects, followed by NRAS in 20% of the population and NF1 in 16%. Figure 9E shows the distribution of mutations in the three major signatures. The extracted signatures were compared to previously validated signatures. The first and second signatures in the detection population were most similar to the UV signature, and the third signature was most closely related to the signature of unknown etiology. As shown in Figure 9E, the mutation signatures found in the detection population are most strongly related to UV-induced DNA damage.

[0125] FIG. 9F shows a bar graph identifying the distribution of subjects corresponding to different levels of responsiveness to immunotherapy for each driver mutation associated with a particular tumor (e.g., BRAF, NRAS). For example, responders were defined as complete response (CR) or partial response (PR). Non-responders were defined as stable disease (SD) or progressive disease (PD). Driver mutations can refer to genetic changes that confer a fundamental growth advantage to cancer cells for neoplastic transformation. In FIG. 9F, subjects with BRAF-mutated tumors tended to respond more aggressively to treatment (47 subjects; exact binomial test; P=0.0258). Response rates for various genomic subtypes did not change significantly from expected response rates. The increase in the number of subjects with progressive disease in the WT gene is likely due to the decrease in the frequency of BRAF, which is typically observed at a higher rate.

[0126] 2. Tumor mutation burden FIG. 10 shows a set of box plots 1000 identifying tumor mutation burden across various driver mutations, disease sites, and subject groups. Box plot 1000 includes box plots 1002, 1004, and 1006. The box plots encompass the interquartile range from the 25th percentile at its lower limit to the 75th percentile at its upper limit, with the median indicated by a horizontal line. The upper whiskers include the maximum value within the 1.5-fold interquartile range above the 75th percentile. The lower whiskers include the minimum value within the 1.5-fold interquartile range below the 25th percentile. The values ​​corresponding to the tumor mutation burden are expressed as log 10 Plotted on a scale.

[0127] Box plot 1002 identifies the tumor mutation burden for each driver mutation. The tumor mutation burden varied significantly between tumors with different driver mutations (Kruskal-Wallis, P=0.00012). Box plot 1004 identifies the tumor mutation burden for each specific site of melanoma disease origin. Box plot 1004 shows the significant overall variation in tumor mutation burden across different sites of disease origin, which is highlighted by contrasting melanomas originating from the head and neck (Kruskal-Wallis, P=0.016).

[0128] Box plot 1006 identifies the tumor mutation burden of a first group of subjects who responded to immunotherapy and a second group of subjects who did not respond to immunotherapy. Comparison of tumor mutation burden in responders and non-responders revealed a significant association (MMW; P=0.049). However, the relatively small difference between tumor mutation burden in responders and non-responders in this population may be due to the confounding effect of melanoma subtype and changes in tumor purity, as these measurements have recently been shown to limit the effectiveness of tumor mutation burden as a predictive biomarker. Thus, tumor mutation burden alone may not predict responsiveness to immunotherapy with high accuracy.

[0129] E. Composite Biomarker Score As described herein, in the embodiments of the present disclosure, it is found that changes in antigen presentation machinery can interfere with neoantigen presentation. These changes have been individually noted to affect the subject's responsiveness to immune checkpoint blockade, and by taking this data into account, it may be possible to improve the performance of predicting responsiveness to immunotherapy. Thus, the composite biomarker score adjusts the neoantigen burden score to take into account subject-specific tumor alterations that may interfere with neoantigen presentation, including HLA mutations, loss of HLA heterozygosity, and B2M mutations. 1. Detection population

[0130] 11A-D show statistical data identifying a composite biomarker score across different subjects, where the composite biomarker score shows improved performance in predicting responsiveness of subjects treated with immunotherapy. In particular, FIG. 11A-D show that the composite biomarker score is more strongly associated with responsiveness to immunotherapy than neoantigen load alone. For example, FIG. 11A shows a set of box plots corresponding to a comparison of the composite biomarker scores between a first group of subjects who responded to immunotherapy and a second group of subjects who did not respond to immunotherapy. As shown in FIG. 11A, the composite biomarker score is significantly higher in responding subjects compared to non-responding subjects (51 subjects; MWW; P=0.002). Thus, the composite biomarker score improved prediction of treatment outcome when compared to neoantigen load. FIG. 11B shows a set of box plots corresponding to a comparison of the composite biomarker scores of groups of subjects (e.g., responding subjects, non-responding subjects) within the validation population. Similar results can be seen in the data from the validation population in FIG. 11B, where subjects in the responding group exhibited significantly higher composite biomarker scores than non-responding subjects (110 subjects; MWW; P=0.010). Referring to FIG. 11A-B, the corresponding box plots encompass the interquartile range from the 25th percentile at its lower limit to the 75th percentile at its upper limit, with the median indicated by a horizontal line. The upper whiskers include the maximum value within the 1.5-fold interquartile range above the 75th percentile. The lower whiskers include the minimum value within the 1.5-fold interquartile range below the 25th percentile.

[0131] Figure 11C shows a line graph identifying a comparison of progression-free survival probability between a first group identified as having a high composite biomarker score and a second group identified as having a low composite biomarker score. As shown in Figure 11C, significantly longer progression-free survival was observed in subjects with high composite biomarker scores compared to subjects with low composite biomarker scores (2-sided KM log-rank test; P=0.0016).

[0132] Figure 11D shows the receiver operating characteristic curves identifying the performance level of the composite biomarker score model. As shown in Figure 11D, the composite biomarker model performed better than the neoantigen burden model, with the area under the curve value of the composite biomarker score increasing from 0.71 to 0.76 and the cross-validation area under the curve value (mean) increasing from 0.69 to 0.75 (log likelihood ratio P=0.0057).

[0133] 2. Test population 12A-B show statistical data identifying the composite biomarker score across various subjects, where the composite biomarker score shows improved performance in predicting progression-free rate and overall survival rate of subjects in the population. In particular, the improvement in the performance level of the composite biomarker score was more pronounced in the validation population. FIG. 12A shows a line graph identifying the comparison of progression-free survival probability between groups of subjects in the validation population, and FIG. 12B shows a line graph identifying the comparison of overall survival rate between groups of subjects in the validation population. In contrast to the results found for neoantigen burden scores in the validation population, FIG. 12A shows that the progression-free survival time of subjects associated with high composite biomarker scores was significantly longer than that of subjects associated with low composite biomarker scores (two-sided KM log-rank test, P=0.05). Furthermore, as shown in FIG. 12B, greater significance was also achieved when analyzing overall survival rate, where the overall survival rate was significantly longer in subjects associated with high composite biomarker scores (two-sided KM log-rank test, P=0.002). The improvement with the combined biomarker score can be biologically interpreted by the finding that 23.5% of subjects in the detection population and 17.27% of subjects in the validation population had at least one mechanism potentially affecting antigen presentation, suggesting that these features may frequently affect the immune system's responsiveness to immunotherapy. 3. HLA gene mutations that affect the combined biomarker score

[0134] Figures 13A-B show statistical data identifying somatic mutations to HLA genes that may contribute to a decreased probability of neoantigen presentation. A review of HLA mutations, particularly those damaging across the entire detection population, revealed deleterious mutations in many subjects. For example, Figure 13A shows examples of somatic mutations identified within samples from the detection population. As shown in Figure 13A, two separate somatic HLA mutations were found in subject 25, including a stop-gain mutation in HLA-A02:01 and a splice region mutation in HLA-B15:01 (allele fractions of 0.473 and 0.368, respectively). These somatic mutations may result in loss of surface expression of HLA-A02:01 and misfolding of HLA-B15:01. A deleterious frameshift mutation was detected within β2-microglobulin (B2M) in subject 38, potentially compromising all MHC class I presentation for that subject.

[0135] FIG. 13B shows a bar graph identifying the relative frequency of neoantigens presented by each HLA gene in subject 25 of the detection population. In FIG. 13B, 38.9% of neoantigens in subject 25 (19.1% for A02:01; 19.8% for B15:01) were predicted to bind to impaired HLA alleles, suggesting a potentially severe impairment of neoantigen presentation. Notably, subject 25 is an outlier with a much higher neoantigen load among non-responders, suggesting that impaired neoantigen presentation beyond that captured by the combined biomarker score may be a contributing factor to immune checkpoint blockade resistance. In another outlier, subject 38 (a non-responder with high neoantigen load), a deleterious frameshift mutation was detected in a high allele fraction in B2M, which may also affect antigen presentation.

[0136] 4. Loss of HLA heterozygosity HLA loss of heterozygosity was further investigated in this population as this loss may affect neoantigen presentation. HLA loss of heterozygosity refers to an acquired resistance mechanism that reduces the ability of tumor neoantigens to be presented to the immune system, promoting immune escape. Since the HLA loss process is governed by selective pressures within the tumor microenvironment, especially at later stages of tumor evolution, it was hypothesized that in the population of late-stage melanoma subjects, allele-specific HLA loss of heterozygosity may contribute to a reduced therapeutic response despite an apparent elevated neoantigen load.

[0137] Loss of heterozygosity of HLA was the most common form of HLA disruption, occurring in 19.6% (10 / 51) of evaluable subjects, with three individuals found to show loss of heterozygosity across all non-homozygous HLAs. Figures 14A-B show examples of sets of decision items specifying a comparison of HLA sequences between normal and corresponding tumor samples of a particular subject. For example, Figure 14A shows a set of decision items specifying a comparison of HLA-A sequences between normal and tumor samples of a subject. Figure 14B shows a set of decision items specifying a comparison of HLA-C sequences between normal and tumor samples of a subject.

[0138] Figure 14A-B provides NGS sequence-based evidence for HLA loss of heterozygosity at HLA-A and HLA-C for 54 subjects in the detection population. HLA-B is not shown. The first row shows the raw read coverage of both homologous alleles in the normal samples. The second row shows the raw read coverage of both homologous alleles in the tumor samples. Both plots include a grey vertical line representing the location of the difference between the two alleles. Due to the strict mapping parameters that require all reads to be mapped without discrepancies, the difference in coverage at the grey line will represent the true difference in coverage between the alleles. The third measure shows the b allele frequency from the normal (grey) and tumor (black) samples. The b allele frequency in the tumor samples must be considered in the context of the b allele frequency in the normal samples due to differences in primer hybridization between the alleles. The fourth measure shows the ratio of coverage between the tumor and normal samples for each allele. These values ​​are normalized by the read depth of the tumor and normal samples across the exome. The expected value for no copy number change is 1 and is shown by the grey dashed line. Both the third and fourth categories show data only for positions with mismatches between the two alleles.

[0139] As shown in Figures 14A-B, matched normal tissue from the subject typically presents allele-specific coverage across HLA genes A and C. In contrast, tumor tissue from this subject shows extensive imbalance in allele-specific coverage across most of the HLAs, with low levels of coverage at HLA-A01:01 and HLA-C07:01. B allele frequencies (b allele frequencies) show absolute differences from normal samples. Consistently low coverage is observed for the missing allele (fourth row in Figures 14A-B), which is predicted to present approximately 54% of the neoantigens in this subject, likely reducing its ability to present to the immune system. VII. Process for Generating Composite Biomarker Scores

[0140] FIG. 15 includes a flow chart 1500 illustrating an example of a method for generating a composite biomarker score, according to some embodiments. The operations described in the flow chart 1500 may be performed by a computer system that performs one or more operations for generating a composite biomarker score based on, for example, a transcriptomic and genomic index. Although the flow chart 1500 describes the operations as a sequential process, in some embodiments, many of the operations may be performed in parallel or simultaneously. Also, the order of the operations may be changed. The operations may include additional steps not shown in the figure. Furthermore, the embodiment of the method may be performed by hardware, software, firmware, middleware, microcode, hardware description languages, or any combination thereof. When performed by software, firmware, middleware, or microcode, the program code or code segments for performing the associated tasks may be stored in a computer readable medium, such as a storage medium.

[0141] In operation 1510, the immunogenomic analysis system processes a biological sample of the subject to obtain genomic and transcriptomic data generated. In some cases, the biological sample includes one or more cancer cells. The genomic data can identify one or more DNA sequences in the biological sample, where whole exome sequencing can be performed to identify one or more DNA sequences. The transcriptomic data can identify one or more RNA sequences in the biological sample, where transcriptomic sequencing can be used to identify one or more RNA sequences. Additionally or alternatively, the genomic and transcriptomic data can be generated from a sample pair including the biological sample of the subject and a reference biological sample, where the reference biological sample does not include one or more cancer cells.

[0142] In operation 1520, the immunogenomic analysis system processes the genomic data to generate a set of genomic indices. Each of the set of genomic indices can represent one or more characteristics corresponding to the corresponding DNA sequence as one or more DNA sequences. In some cases, the set of genomic indices includes (i) a quantitative or classifying indices representing one or more characteristics for each of one or more somatic mutations in the one or more DNA sequences; (ii) a classifying indices indicating whether loss of heterozygosity has occurred in at least one human leukocyte antigen (HLA) gene of the biological sample; and (iii) a quantitative or classifying indices representing a predicted tumor mutation load. For HLA loss of heterozygosity, the corresponding classifying indices can be generated by fitting the genomic data to an HLA deletion discrimination machine learning model.

[0143] In operation 1530, the immunogenomic analysis system processes the transcriptome data to generate a set of transcriptome indices. Each of the set of transcriptome indices can represent one or more characteristics corresponding to a set of peptides translated from a corresponding RNA sequence of the one or more RNA sequences. In some cases, the set of transcriptome indices includes: (i) a quantitative indices or classifiers representing a predicted neoantigen load of the biological sample; (ii) a quantitative indices or classifiers representing one or more characteristics of each of the one or more candidate neoantigens detected in the biological sample; (iii) a quantitative indices or classifiers representing one or more characteristics of each of the one or more HLA proteins for which loss of cell surface presentation is detected; (iv) a quantitative indices or classifiers representing one or more characteristics corresponding to HLA genes encoding the one or more HLA proteins for which loss of cell surface presentation is detected; (v) a quantitative indices or classifiers representing expression levels of sequences corresponding to immune cells; and (vi) a quantitative indices or classifiers representing expression levels of one or more T cell receptors detected in the biological sample. For HLA proteins where loss of cell surface presentation is detected, corresponding indices can be generated by fitting genomic and transcriptomic data to neoantigen presentation predictive machine learning models.

[0144] In operation 1540, the immunogenomic analysis system generates a composite biomarker score derived from the set of genomic indices and the set of transcriptomic indices. In some cases, the immunogenomic analysis system generates the composite biomarker score by (i) weighting each genomic indices in the set of genomic indices with a weight value determined based on a corresponding transcriptomic indices in the set of transcriptomic indices, and (ii) generating the composite biomarker score using the weighted genomic indices.

[0145] In act 1550, the immunogenomic analysis system determines a predicted level of the subject's responsiveness to a particular type of immunotherapy treatment based on the composite biomarker score.

[0146] In operation 1560, the immunogenomic analysis system outputs a result corresponding to the predicted level of responsiveness of the subject. The result may be a report that identifies (i) a recommendation of a particular treatment; (ii) a recommendation to administer the particular treatment to the human subject; and / or (iii) a recommendation not to administer the particular treatment to the human subject based on the predicted level of responsiveness of the subject to the particular treatment. In some embodiments, the recommended treatment is administered to the human subject. Process 1500 then ends.

[0147] VIII. Other Considerations Although the subject matter has been described in detail with respect to certain embodiments thereof, it will be appreciated that those skilled in the art, upon understanding the foregoing, can easily generate modifications, variations, and equivalents to the embodiments. It will be appreciated, therefore, that the present disclosure is intended to be presented by way of example and not by way of limitation, and is not intended to exclude modifications, variations, and / or additions to the subject matter that are readily apparent to those skilled in the art. In fact, the methods and systems described herein may be embodied in a variety of other forms, and various omissions, substitutions, and changes in the forms of the methods and systems described herein may be made without departing from the spirit of the disclosure. The appended claims and their equivalents are intended to cover such forms or modifications that fall within the scope and spirit of the present disclosure.

[0148] Unless otherwise indicated, descriptions utilizing terms such as "processing," "computing," "calculating," "determining," and "identifying" throughout this specification are understood to refer to operations or processes of a computing device, such as one or more computers or similar electronic computing devices, that manipulate or transform data represented as physical electronic or magnetic quantities in memories, registers, or other information storage, transmission, or display devices of a computing platform.

[0149] The one or more systems described herein are not limited to a particular hardware construct or configuration. A computing device may include any suitable arrangement of components that provide a result conditioned on one or more inputs. Suitable computing devices include general-purpose microprocessor-based computing systems that utilize stored software to program or configure computing systems ranging from general-purpose computing devices to special-purpose computing devices that execute one or more embodiments of the present subject matter. Any suitable programming, scripting, or other form of language or combination of languages ​​may be used to implement the teachings contained herein in software that can be used to program or configure a computing device.

[0150] In some embodiments, the embodiments may be implemented by the operation of such a computing device. The order of the blocks presented in the above examples may be changed, e.g., the blocks may be rearranged, combined, and / or divided into sub-blocks. Certain blocks or processes may be performed in parallel.

[0151] As used herein, conditional terms, particularly "can," "could," "might," "may," "eg," and the like, are intended to convey that, unless otherwise specified or interpreted otherwise within the context in which they are used, a particular example generally includes certain features, elements, and / or steps, and other examples do not include them. Thus, such conditional terms are not intended to imply that the features, elements, and / or steps are generally required in any way for one or more examples, or that one or more examples necessarily include logic, whether or not input or written by the author, that determines whether those features, elements, and / or steps are included or performed in any particular example.

[0152] Terms such as "including," "including," "having," and the like are synonymous and are used inclusively in a non-limiting manner and do not exclude additional elements, features, actions, operations, etc. Additionally, the term "or," for example, when used to connect a list of elements, is used in an inclusive (and not exclusive) sense to mean one, some, or all of the elements in the list. Use of "adapted to" or "configured to" herein refers to phrases that do not exclude apparatus adapted or configured to perform additional operations or steps. Additionally, use of "based on" means inclusive and non-limiting such that a process, step, calculation, or other operation "based on" one or more stated conditions or values ​​may, in fact, be based on additional conditions or values ​​beyond those stated. Similarly, the use of "based at least in part on" is meant to be open-ended and inclusive, such that a process, step, calculation, or other operation "based at least in part on" one or more recited conditions or values ​​may, in fact, be based on additional conditions or values ​​beyond the recited conditions or values. The headings, lists, and numbering contained herein are for ease of description only and are not meant to be limiting.

[0153] The various features and processes described above may be used independently of each other or may be combined in various ways. All possible combinations and subcombinations are intended to fall within the scope of the present disclosure. Furthermore, certain method or process blocks may be omitted in some implementations. The methods and processes described herein are also not limited to any particular order, and the blocks or states associated therewith may be performed in other orders as appropriate. For example, the described blocks or states may be performed in orders other than those specifically disclosed, or multiple blocks or states may be combined into a single block or state. The example blocks or states may be performed sequentially, in parallel, or in other manners. Blocks or states may be added to or deleted from the disclosed examples. Similarly, the example systems and components described herein may be configured differently than described. For example, elements may be added to, deleted from, or reorganized relative to the disclosed examples.

Claims

1. obtaining genomic data identifying one or more DNA sequences in the biological sample and transcriptomic data identifying one or more RNA sequences in the biological sample generated by processing a biological sample from a subject comprising one or more cancer cells; generating a set of genomic indicators based on the genomic data, each of the set of genomic indicators representing one or more DNA sequences corresponding to one or more characteristics of the corresponding DNA sequence; generating a set of transcriptomic signatures based on the transcriptomic data, each of the set of transcriptomic signatures representing one or more features corresponding to a set of peptides translated from corresponding RNA sequences of the one or more RNA sequences; identifying a composite biomarker score derived from said set of genomic features and said set of transcriptomic features; determining a predictive level of the subject's responsiveness to a particular type of immunotherapy treatment based on the composite biomarker score; and outputting a result corresponding to the predicted level of responsiveness of the subject; The method includes:

2. 2. The method of claim 1, wherein generating the set of genomic features comprises determining one or more characteristic quantitative or classifiers for each of one or more somatic mutations in one or more DNA sequences.

3. 3. The method of claim 1 or 2, wherein generating the set of genomic indicators comprises determining a classification indicator indicative of whether loss of heterozygosity has occurred within at least one human leukocyte antigen (HLA) gene of the biological sample.

4. 4. The method of claim 3, wherein determining an indication of whether loss of heterozygosity has occurred comprises fitting the genomic data to an HLA deletion discrimination machine learning model to generate an output corresponding to the indication of whether loss of heterozygosity has occurred.

5. The method of any one of claims 1 to 4, wherein the step of generating a set of transcriptomic signatures comprises determining a quantitative or classifying signature representative of the predicted neoantigen load of the biological sample.

6. 6. The method of claim 1, wherein the step of generating a set of transcriptomic signatures comprises determining, based on the genomic data and the transcriptomic data, quantitative or classifying signatures representative of one or more characteristics of each of the one or more candidate neoantigens detected in the biological sample.

7. 7. The method of any one of claims 1 to 6, wherein the step of generating a set of transcriptomic signatures comprises generating a quantitative or classification signature representative of one or more characteristics of each of the one or more HLA proteins for which loss of cell surface presentation is detected.

8. 8. The method of claim 7, wherein the step of generating a set of transcriptomic signatures comprises generating, based on the transcriptomic data, a quantitative or classifier signature representative of one or more characteristics corresponding to HLA genes encoding one or more HLA proteins for which loss of cell surface presentation is detected.

9. 8. The method of claim 7, wherein generating the set of transcriptomic indices comprises fitting the genomic data and the transcriptomic data to a neoantigen presentation predictive machine learning model to generate the quantitative or classifier indices representative of one or more characteristics of each of the one or more HLA proteins.

10. 10. The method of any one of claims 1 to 9, wherein the step of generating a set of transcriptomic signatures comprises determining a quantitative or classifier signature representative of the expression level of one or more T cell receptors detected in the biological sample.

11. The method of any one of claims 1 to 10, wherein the biological sample is taken from a tumor of the subject, and generating the set of transcriptomic indices comprises determining quantitative or classifier indices representative of expression levels of sequences corresponding to immune cells.

12. The method of any one of claims 1 to 11, wherein the step of obtaining the genomic and transcriptomic data comprises identifying the one or more DNA sequences using whole exome sequencing.

13. The method of any one of claims 1 to 12, wherein the step of obtaining the genomic data and the transcriptomic data comprises identifying the one or more RNA sequences using transcriptome sequencing.

14. 14. The method of any one of claims 1 to 13, wherein obtaining the genomic data and the transcriptomic data comprises generating the genomic data and the transcriptomic data from the biological sample of the subject and a reference biological sample, wherein the reference biological sample does not contain one or more cancer cells.

15. The step of generating the composite biomarker score comprises: weighting each genomic indicator in the set of genomic indicators with a weight value determined based on a corresponding transcriptomic indicator in the set of transcriptomic indicators; and The method of any one of claims 1 to 14, comprising using the weighted genomic index to generate the composite biomarker score.

16. one or more data processors; and A system comprising a non-transitory computer-readable storage medium comprising instructions that, when executed on the one or more data processors, cause the one or more data processors to perform some or all of one or more methods disclosed herein.

17. A computer program product tangibly embodied in a non-transitory machine-readable storage medium comprising instructions configured to cause one or more data processors to perform some or all of one or more of the methods disclosed herein.

Images