Immuno-oncology applications using next generation sequencing

Abstract

Provided herein are systems and methods for generating an immune-oncology profile from a biological sample. The immune-oncology profile can include the proportion or percentage of immune cells, expression of immune escape genes, and/or mutational burden. The immune-oncology profile may allow the generation of classifiers for making prognostic or diagnostic predictions.

What is claimed is: 
     
         1 . A computer-implemented method for analyzing a biological sample obtained from a subject having or suspected of having a disease or condition, comprising:
 (a) obtaining gene expression data comprising (i) an expression level of at least one immune modulatory gene and (ii) expression levels of a plurality of expression signature genes from the biological sample;   (b) using a deconvolution algorithm to process said expression levels of said plurality of expression signature genes to identify and quantify a percentage of at least one cell type that is present in the biological sample; and   (c) using a classifier to analyze the expression level of the at least one immune modulatory gene and the percentage of the at least one cell type from (b) to determine a likelihood that said subject will be responsive or non-responsive to therapy.   
     
     
         2 . The method of  claim 1 , wherein the therapy comprises immunotherapy. 
     
     
         3 . The method of  claim 1 , further comprising providing instructions to start, stop, change, or continue the therapy. 
     
     
         4 . The method of  claim 1 , wherein the disease or condition is cancer, and wherein (c) comprises determining that said likelihood that said subject will be responsive or non-responsive to said therapy for said cancer. 
     
     
         5 . The method of  claim 1 , wherein the at least one cell type comprises at least one immune cell type. 
     
     
         6 . The method of  claim 5 , wherein the at least one immune cell type comprises M1 macrophages, M2 macrophages, CD19+ B cells, CD14+ monocytes, CD56+ NK cells, CD8+ T cells, Treg cells, CD4+ T cells, or any combination thereof. 
     
     
         7 . The method of  claim 1 , wherein the at least one immune modulatory gene comprises CTLA4, OX40, PD-1, IDO1, CD47, PD-L1, TIM-3, BTLA, ICOS, ARG1, or any combination thereof. 
     
     
         8 . The method of  claim 1 , wherein the classifier is generated using a machine learning algorithm. 
     
     
         9 . The method of  claim 8 , wherein the machine learning algorithm is a random forest algorithm. 
     
     
         10 . The method of  claim 1 , wherein the deconvolution algorithm in (b) applies a deconvolution matrix to process said expression levels of said plurality of expression signature genes to identify and quantify the percentage of the at least one cell type. 
     
     
         11 . The method of  claim 10 , wherein the deconvolution matrix comprises a plurality of immune cell expression signature genes. 
     
     
         12 . The method of  claim 10 , wherein the deconvolution matrix comprises a plurality of tumor cell expression signature genes. 
     
     
         13 . The method of  claim 10 , wherein the deconvolution matrix comprises a plurality of cell types, each cell type comprising a plurality of expression signature genes, wherein expression count for each expression signature gene is normalized across the plurality of cell types. 
     
     
         14 . The method of  claim 1 , wherein the deconvolution algorithm processes said expression levels of said plurality of expression signature genes using linear least-squares regression (LLSR), quadratic programming (QP), perturbation model for gene expression deconvolution (PERT), robust linear regression (RLR), microarray microdissection with analysis of differences (MMAD), digital sorting algorithm (DSA), or support vector regression. 
     
     
         15 . The method of  claim 14 , wherein the deconvolution algorithm performs an RNA normalization step to compensate for variation in RNA quantity amongst the at least one cell type in order to improve accuracy of the percentage of the at least one cell type. 
     
     
         16 . The method of  claim 15 , wherein the deconvolution algorithm is a machine learning algorithm trained using comparison data comprising an actual percentage of the at least one cell type. 
     
     
         17 . The method of  claim 1 , wherein the gene expression data and the plurality of expression signature genes are obtained from the biological sample using next generation RNA sequencing. 
     
     
         18 . The method of  claim 1 , further comprising processing the gene expression data to determine mutational burden for the biological sample and inputting the mutational burden into the classifier for analysis in order to enhance classification of the biological sample. 
     
     
         19 . The method of  claim 1 , wherein the classifier is trained on data from no more than 50 samples and provides an accuracy of at least 85%. 
     
     
         20 . A system comprising for analyzing a biological sample obtained from a subject having or suspected of having a disease or condition, comprising:
 a database comprising gene expression data comprising (i) an expression level of at least one immune modulatory gene and (ii) expression levels of a plurality of expression signature genes from the biological sample; and   at least one computer processor that is coupled to said database, wherein said at least one computer processor is programmed to:   (a) use a deconvolution algorithm to process said expression levels of said plurality of expression signature genes to identify and quantify a percentage of at least one cell type that is present in the biological sample;   (b) use a classifier to analyze the expression level of the at least one immune modulatory gene and the percentage of the at least one cell type from (b) to determine a likelihood that said subject will be responsive or non-responsive to therapy.   
     
     
         21 . The system of  claim 20 , wherein the at least one cell type comprises M1 macrophages, M2 macrophages, CD19+ B cells, CD14+ monocytes, CD56+ NK cells, CD8+ T cells, Treg cells, CD4+ T cells, or any combination thereof. 
     
     
         22 . The system of  claim 20 , wherein the at least one immune modulatory gene comprises CTLA4, OX40, PD-1, IDO1, CD47, PD-L1, TIM-3, BTLA, ICOS, ARG1, or any combination thereof. 
     
     
         23 . The system of  claim 20 , wherein the classifier is generated using a machine learning algorithm. 
     
     
         24 . The system of  claim 23 , wherein the machine learning algorithm is a random forest algorithm. 
     
     
         25 . The system of  claim 20 , wherein the deconvolution algorithm in (a) applies a deconvolution matrix to process said expression levels of said plurality of expression signature genes to identify and quantify the percentage of the at least one cell type. 
     
     
         26 . The system of  claim 25 , wherein the deconvolution algorithm performs an RNA normalization step to compensate for variation in RNA quantity amongst the at least one cell type in order to improve accuracy of the percentage of the at least one cell type. 
     
     
         27 . The system of  claim 20 , wherein the gene expression data and the plurality of expression signature genes are obtained from the biological sample using next generation RNA sequencing. 
     
     
         28 . The system of  claim 27 , wherein the at least one processor is further programmed to obtain mutational burden data for the biological sample and inputting the mutational burden data into the classifier for analysis in order to enhance classification of the biological sample. 
     
     
         29 . The system of any one of  claims 20 , wherein the classifier is trained on data from no more than 50 samples and provides an accuracy of at least 85%. 
     
     
         30 . Non-transitory computer readable medium comprising machine executable code that, upon execution by one or more computer processors, performs a method comprising:
 (a) obtaining gene expression data comprising (i) an expression level of at least one immune modulatory gene and (ii) expression levels of a plurality of expression signature genes from the biological sample;   (b) using a deconvolution algorithm to process said expression levels of said plurality of expression signature genes to identify and quantify a percentage of at least one cell type that is present in the biological sample; and   (c) using a classifier to analyze the expression level of the at least one immune modulatory gene and the percentage of the at least one cell type from (b) to determine a likelihood that said subject will be responsive or non-responsive to therapy.