The repertoires of tumor-infiltrating T cells and B cells are rich sources of information about cancer-immune interactions and provide insights on cancer immunotherapy targets. Efforts have been made to characterize B/T cell repertoires in solid tumors using cell sorting followed by targeted deep sequencing. However, these approaches may produce biased estimates during tissue disaggregation and can be expensive when applied to large sample cohorts. Massively parallel mRNA sequencing (RNA-seq) technology has become the mainstream method to profile gene expression and thousands of solid tumor RNA-seq profiles are available in the public domain. The rich collection of tumor RNA-seq datasets provides an alternative approach to study tumor-infiltrating B/T cell repertoires in solid tumors. Our team has recently developed a statistical method TIMER for deconvolving different immune components in the tumor microenvironment, and TRUST for inferring the hypervariable complementarity determining regions (CDRs) of the tumor infiltrating T cell receptor (TCR) repertoire from bulk tumor RNA-seq data in the public domain. Our preliminary analysis indicated that there are approximately ten times as many B cell receptor (BCR) reads and TCR reads, suggesting that extracting the BCR repertoires from bulk tumor RNA-seq could reveal important insights on B cell mediated tumor immunity.
The aims of this proposal are: to extend our TRUST algorithm to extract B cell receptor (BCR) repertoires from tumor RNA-seq data, and identify somatic hypermutations and immunoglobin class switches (Aim 1); to systematically analyze TCR and BCR repertoires from large scale tumor RNA-seq cohorts, and develop a user friendly web interface to allow cancer immunologists or immuno-oncologists to investigate tumor-immune associations (Aim 2); to promote the utility of our tumor immune resource through collaborations, cloud sharing, and outreach (Aim 3). We will deliver a robust bioinformatics algorithm to systematically identify BCR / TCR repertoires from bulk tumor RNA-seq data and a user-friendly resource for cancer immunologists or immuno-oncologists to explore tumor- immune interactions from large tumor profiling cohorts in the public as well as their unpublished data. The successful execution of this proposal has the potential to inform clinical practice of cancer immunotherapies, including adoptive T cell transfer, therapeutic cancer vaccines or antibodies. Our proposed cancer immunology algorithm and resource will be a unique addition to the array of bioinformatics tools developed by the Information Technology for Cancer Research at the National Cancer Institute.
Cancer immunotherapies, which depend on boosting the patient immune system to kill cancer cells, have recently been remarkably successful in the clinic. We propose to develop the computational algorithms and resources that enable cancer immunologists and immuno-oncologists to investigate T cell and B cell related immunity in tumors by better utilizing existing large-scale tumor profiling data.