The paper introduces PROGENy (Pathway RespOnsive GENes), a novel computational method developed to accurately infer signaling pathway activity in cancer from gene expression data. The authors highlight that traditional methods of mapping gene expression to pathway components are often inadequate because they fail to account for post-translational modifications and are often too context-specific. PROGENy overcomes these limitations by leveraging a large collection of public perturbation experiments to identify a common core of consistently deregulated genes that respond to pathway stimuli. The article demonstrates that PROGENy significantly outperforms existing pathway analysis methods in recovering the effects of known driver mutations, predicting drug sensitivity in cell lines, and distinguishing between oncogenic and tumor-suppressor pathways in terms of patient survival. Ultimately, the research supports the use of a downstream gene expression readout rather than direct transcript mapping for inferring pathway activity.

References:

• Schubert M, Klinger B, Klünemann M, et al. Perturbation-response genes reveal signaling footprints in cancer gene expression[J]. Nature communications, 2018, 9(1): 20.