Multi-omic integration with cosmosR and ocEAn to study cross-talks between signalling and metabolism in diseases.
Aurelien Dugourd,Julio Saez-Rodriguez
Heidelberg Medical University
Abstract
COSMOS (Causal Oriented Search of Multi-Omic Space) is a method that integrates multi-omic data such as (but not limited to) phosphoproteomics, transcriptomics, and metabolomics data sets. COSMOS leverages extensive prior knowledge of signaling pathways, metabolic networks, and gene regulation with computational methods to estimate activities of transcription factors and kinases as well as network-level causal reasoning. This pipeline can provide mechanistic explanations for experimental observations across multiple omic data sets. Recently, we have expanded and further curated the metabolic reaction prior knowledge of COSMOS, allowing to double the metabolic coverage of COSMOS. We have simplified the required input to run COSMOS, and added functions to make it easier to generate such inputs. We have improved the pipeline to make it more efficient, allowing for the use of free solvers such as CBC. To demonstrate the flexibility of COSMOS, we have formatted multi-omic datasets from 58 cell lines generated by the NCI-60 Human Tumor Cell Lines Screen into sets of COSMOS-ready input, allowing users to play around a wide variety of cell line datasets to try COSMOS. Finally, supported by COSMOS reaction network, we have developed a new R package called ocEAn to study patterns of metabolic deregulations around metabolic enzymes. ocEAn defines metabolic enzyme footprint from a reaction network and use them to explore coordinated deregulations of metabolite abundances with respect to their position relative to metabolic enzymes. Such coordinated deregulations can help us identify metabolic enzymes that act as bottle neck or metabolic syphons in diseases.