An unsupervised feature extraction and selection strategy for identifying epithelial-mesenchymal transition state metrics in breast cancer and melanoma

Abstract

Digital cytometry is opening up new avenues to better understand the heterogeneous cell types present within the tumor microenvironment. While the focus is towards elucidating immune and stromal cells as clinical correlates, there is still a need to better understand how a change in tumor cell phenotype, such as the epithelial-mesenchymal transition, influences the immune contexture. To complement existing digital cytometry methods, our objective was to develop an unsupervised gene signature capturing a change in differentiation state that is tailored to the specific cellular context of breast cancer and melanoma, as a illustrative example. Towards this aim, we used principal component analysis coupled with resampling to develop unsupervised gene expression-based state metrics specific for the cellular context that characterize the state of cellular differentiation within an epithelial to mesenchymal-like state space and independently correlate with metastatic potential. First developed using cell line data, the orthogonal state metrics were refined to exclude the contributions of normal fibroblasts and to provide tissue-level state estimates based on bulk tissue RNA-seq measures. The resulting gene expression-based metrics for differentiation state aim to inform a more holistic view of how the malignant cell phenotype influences the immune contexture within the tumor microenvironment.