SignalingProfiler 2.0: a network-based approach to bridge multi-omics data to phenotypic hallmarks

Abstract

Unraveling the cellular signaling remodeling upon a perturbation is a fundamental challenge to understand disease mechanisms and to identify potential drug targets. In this pursuit, computational tools that generate mechanistic hypotheses from multi-omics data have invaluable potential. Here we present SignalingProfiler 2.0, a multi-step pipeline to systematically derive context-specific signaling models by integrating proteogenomic data with prior knowledge-causal networks. This is a freely accessible and flexible tool that incorporates statistical, footprint-based, and graph algorithms to accelerate the integration and interpretation of multi-omics data. Through benchmarking and rigorous parameter selection on a proof-of-concept study, performed in metformin-treated breast cancer cells, we demonstrate the tools ability to generate a hierarchical mechanistic network that recapitulates novel and known drug-perturbed signaling and phenotypic outcomes. In summary, SignalingProfiler 2.0 addresses the emergent need to derive biologically relevant information from complex multi-omics data by extracting interpretable networks.