The M2 Pipeline: A Novel Tool for Rapidly Analyzing Host Metabolic Profiles from RNA-seq Data
Abstract— A novel bioinformatics and data analytics pipeline that utilizes global gene expression datasets in combination with advanced computational modeling of metabolic pathways has been applied to predicting the metabolic profiles during viral infections. The Modeling Metabolism (M2) pipeline was used to analyze and compare complex metabolic profiles of four viral respiratory infections (rhinovirus infection, flu, COVID-19, and MERS-CoV infection) based on host RNA-seq datasets. During infection, multiple factors can affect the metabolic state of the host. Metabolic changes are biomarkers of host antiviral responses and viral pathogenesis. Therefore, analyzing the metabolic response to infectious diseases might provide valuable mechanistic insights into host-pathogen interactions and highlight metabolic changes that play important roles in disease pathogenesis. By using the M2 pipeline, we predicted that influenza virus, SARS-CoV-2, and MERS-CoV infections resulted in increased glycolytic activities and a reduced capacity for oxidative phosphorylation. Furthermore, SARS-CoV-2 and MERS-CoV infections both presented dysregulated pentose phosphate pathways (PPP), while influenza infection elicited an increase in PPP activity in correlation with disease severity and mortality. Notably, rhinovirus infection, the mildest respiratory infection studied, had little effect on the overall perturbation of host cellular metabolism. The M2
pipeline provides a rapid, comprehensive, and systems- wide analysis of metabolic profiles from host response RNA-seq datasets during emerging and re-emerging infections. Keywords—Bioinformatics, Computational Biology, Metabolism, Infectious Disease, Immunology, RNA-seq