Functional characterization of somatic mutations in cancer using network-based inference of protein activity

Area of Science:

• Oncology
• Bioinformatics
• Genomics

Background:

• Personalized cancer treatment requires identifying dysregulated oncoproteins.
• Accurate inference of protein activity from genetic alterations is challenging.
• Direct protein activity measurements are often not feasible.

Purpose of the Study:

• To introduce and validate a novel algorithm, VIPER, for assessing protein activity from gene expression data.
• To evaluate the functional relevance of genetic alterations in regulatory proteins across The Cancer Genome Atlas (TCGA) samples.
• To improve the prediction of targeted therapy sensitivity.

Main Methods:

• Development and experimental validation of the virtual inference of protein activity by enriched regulon analysis (VIPER) algorithm.
• Application of VIPER to gene expression data from The Cancer Genome Atlas (TCGA).
• In vitro assays to compare VIPER-inferred activity with mutational analysis for predicting drug sensitivity.

Main Results:

• VIPER accurately assesses protein activity from gene expression data.
• VIPER identified tumors with aberrant oncoprotein activity independent of mutations.
• VIPER-inferred protein activity was superior to mutational analysis in predicting sensitivity to targeted inhibitors.

Conclusions:

• VIPER provides a robust method for inferring protein activity, aiding in personalized cancer therapy.
• The algorithm can identify actionable oncoprotein dysregulation even in the absence of detectable mutations.
• VIPER enhances the prediction of treatment response, guiding more effective therapeutic strategies.