Prophetic Granger Causality to infer gene regulatory networks

Abstract

We introduce a novel method called Prophetic Granger Causality (PGC) for inferring gene regulatory networks (GRNs) from protein-level time series data. The method uses an L1-penalized regression adaptation of Granger Causality to model protein levels as a function of time, stimuli, and other perturbations. When combined with a data-independent network prior, the framework outperformed all other methods submitted to the HPN-DREAM 8 breast cancer network inference challenge. Our investigations reveal that PGC provides complementary information to other approaches, raising the performance of ensemble learners, while on its own achieves moderate performance. Thus, PGC serves as a valuable new tool in the bioinformatics toolkit for analyzing temporal datasets. We investigate the general and cell-specific interactions predicted by our method and find several novel interactions, demonstrating the utility of the approach in charting new tumor wiring.

Synthesizing Signaling Pathways from Temporal Phosphoproteomic Data

“…TPS does not require perturbations to reconstruct pathways ( Ciaccio et al, 2015 ; Molinelli et al, 2013 ; Terfve et al, 2015 ). Participants in the HPN-DREAM network inference challenge ( Hill et al, 2016 ) inferred signaling networks from time series data for tens of phosphoproteins, but the top methods either did not scale to our dataset (PropheticGranger; Carlin et al, 2017 ) or did not perform well (FunChisq; Zhang and Song, 2013 ). Other algorithms that integrate temporal information with PPI networks ( Budak et al, 2015 ; Gitter and Bar-Joseph, 2013 ; Jain et al, 2016 ; Norman and Cicek, 2018 ; Patil et al, 2013 ) do not evaluate and summarize all pathway models that are supported by the network and phosphorylation timing constraints.…”

Copycat Layout: Network layout alignment via Cytoscape Automation

“…In the case of inferring gene regulatory networks 9 , it is beneficial to show mutations at each time-step without touching the unchanged parts of the network. Identifying and visualizing the changes over time is much easier with Cytoscape Automation.…”

Review of Causal Discovery Methods Based on Graphical Models

“…For instance, the idea of the PC algorithm was adopted to infer causal relationships among phenotypes (Neto et al, 2010), to estimate gene regulatory networks (Zhang et al, 2011b), and to model the isoprenoid gene network in Arabidopsis thaliana (Wille et al, 2004). Granger causal analysis received a number of applications in estimation of gene regulatory networks; see, e.g., (Michailidis and d'Alché Buc, 2013; Emad and Milenkovic, 2014; Carlin et al, 2017; Yang et al, 2017; Finkle et al, 2018), and similarly, some findings were based on dynamic Bayesian network learning from observational biological data (Yu et al, 2004; Wu and Liu, 2008; Vasimuddin and Srinivas, 2017). There are also applications of network inference methods to leverage multiple data sets (Reiss et al, 2006; Joshi et al, 2015; Zitnik and Zupan, 2015; Omranian et al, 2016).…”