Detecting and dissecting signaling crosstalk via the multilayer network integration of signaling and regulatory interactions

Abstract

The versatility of cellular response arises from the communication, or crosstalk, of signaling pathways in a complex network of signaling and transcriptional regulatory interactions. Understanding the various mechanisms underlying crosstalk on a global scale requires untargeted computational approaches. We present a network-based statistical approach, MuXTalk, that uses high-dimensional edges called multilinks to model the unique ways in which signaling and regulatory interactions can interface. We demonstrate that the signaling-regulatory interface is located primarily in the intermediary region between signaling pathways where crosstalk occurs, and that multilinks can differentiate between distinct signaling-transcriptional mechanisms. Using statistically over-represented multilinks as proxies of crosstalk, we predict crosstalk among 60 signaling pathways, expanding currently available crosstalk databases by more than five-fold. MuXTalk surpasses existing methods in terms of prediction performance, identifies additions to manual curation efforts, and pinpoints potential mediators of crosstalk for each prediction. Moreover, it accommodates the inherent context-dependence of crosstalk, allowing future applications to cell type- and disease-specific crosstalk.

Advancements in computational modelling of biological systems: seventh annual SysMod meeting

“…The first session ended with two flash talks. Lavanya Raajaraam discussed the Co-FSEOF framework to study metabolite coproduction ( Raajaraam and Raman 2021 ) and Arda Halu presented MuXTalk, a tool to model signalling crosstalk using multilayer networks ( Halu et al 2022 ).…”