Genome-scale mechanistic modeling of signaling pathways made easy: A bioconductor/cytoscape/web server framework for the analysis of omic data

Abstract

Genome-scale mechanistic models of pathways are gaining importance for genomic data interpretation because they provide a natural link between genotype measurements (transcriptomics or genomics data) and the phenotype of the cell (its functional behavior). Moreover, mechanistic models can be used to predict the potential effect of interventions, including drug inhibitions. Here, we present the implementation of a mechanistic model of cell signaling for the interpretation of transcriptomic data as an R/Bioconductor package, a Cytoscape plugin and a web tool with enhanced functionality which includes building interpretable predictors, estimation of the effect of perturbations and assessment of the effect of mutations in complex scenarios.

Crosstalk between Metabolite Production and Signaling Activity in Breast Cancer

“…The Hipathia mechanistic model, which models the activity of signaling pathways from gene expression data [ 24 ], was used to estimate signaling activity. In particular, a R/Bioconductor implementation (v2.14.0) was used [ 33 ]. Hipathia uses KEGG [ 16 ], signaling pathways, which are decomposed into elementary signaling circuits, which can be considered self-regulating functional units of the cell [ 24 , 33 , 86 ].…”

“…Here, a study on the crosstalk between metabolism and signaling has been carried out in BRCA, using mechanistic models to simultaneously infer from gene expression data the production of metabolites [ 21 , 22 ] and the activity of signaling pathways [ 24 , 33 ]. Importantly, the use of machine learning methods, such as Gaussian Processes (GP) [ 34 ] along with the SHapley Additive exPlanations (SHAP) method [ 35 ] that allows exploring causality, has allowed to relate the production of certain metabolites to the different activation statuses of several signaling circuits.…”

Integrating knowledge and omics to decipher mechanisms via large‐scale models of signaling networks

“…Similarly, the software packages accompanying the methods are different in form, ranging from software with graphical user interface to collections of R and Python scripts. There are multiple publications describing different versions of certain methods, as is the case for CLIPPER (Massa et al , 2010 ; Martini et al , 2013 ), HiPathia (Hidalgo et al , 2017 ; Rian et al , 2021 ) and PHONEMES (Terfve et al , 2015 ; Gjerga et al , 2021 ). Relatedly, there are also articles where the reviewed methods were expanded, refined, or applied to different data modalities, such as TieDIE (Paull et al , 2013 ) to phosphoproteomics (Drake et al , 2016 ), HiPathia (Hidalgo et al , 2017 ) to single‐cell transcriptomics (Falco et al , 2020 ), or CARNIVAL's core formulation (Liu et al , 2019 ) to multi‐omics in COSMOS (Dugourd et al , 2021 ).…”

Discovering potential interactions between rare diseases and COVID-19 by combining mechanistic models of viral infection with statistical modeling

“…The mechanistic model of the COVID-19 disease map allows transforming the transcriptional and/or mutational profiles of the genes involved in the map into signaling activity profiles of the corresponding signaling circuits ( 14 , 22 ). In particular, the hipathia ( 22 , 65 ) algorithm has been used here, which has demonstrated to outperform other similar algorithms ( 46 ). Specifically, the mechanistic model simulates the behavior of the signaling circuits of the COVID-19 disease map.…”