Simultaneous enumeration of cancer and immune cell types from bulk tumor gene expression data

Abstract

Immune cells infiltrating tumors can have important impact on tumor progression and response to therapy. We present an efficient algorithm to simultaneously estimate the fraction of cancer and immune cell types from bulk tumor gene expression data. Our method integrates novel gene expression profiles from each major non-malignant cell type found in tumors, renormalization based on cell-type-specific mRNA content, and the ability to consider uncharacterized and possibly highly variable cell types. Feasibility is demonstrated by validation with flow cytometry, immunohistochemistry and single-cell RNA-Seq analyses of human melanoma and colorectal tumor specimens. Altogether, our work not only improves accuracy but also broadens the scope of absolute cell fraction predictions from tumor gene expression data, and provides a unique novel experimental benchmark for immunogenomics analyses in cancer research ( http://epic.gfellerlab.org).

eLife digest

Malignant tumors do not only contain cancer cells. Normal cells from the body also infiltrate tumors. These often include a variety of immune cells that can help detect and kill cancer cells. Many evidences suggest that the proportion of different immune cell types in a tumor can affect tumor growth and which treatments are effective.

Researchers often study tumors by measuring the expression of genes, i.e., which genes are active in tumors. However, the proportion of different cell types in the tumor is often not measured for tumors studied at the gene expression level.

Racle et al. have now demonstrated that a new computer-based tool can accurately detect all the main cell types in a tumor directly from the expression of genes in this tumor. The tool is called “Estimating the Proportion of Immune and Cancer cells” – or EPIC for short. It compares the level of expression of genes in a tumor with a library of the gene expression profiles from specific cell types that can be found in tumors and uses this information to predict how many of each type of cell are present. Experimental measurements of several human tumors confirmed that EPIC’s predictions are accurate.

EPIC is freely available online. Since the active genes in tumors from many patients have already been documented together with clinical data, researchers could use EPIC to investigate whether the cell types in a tumor affect how harmful it is or how well a particular treatment works on it. In the future, this information could help to identify the best treatment for a particular patient and may reveal new genes that cause malignant tumors to develop and grow.