Personalized Constraint-Based Modeling of Microbial Communities from Metagenomic Data

High-throughput metagenomic sequencing techniques such as 16S rRNA and shotgun sequencing have enabled an unprecedented understanding of the structure and function of microbiome communities such as the human gut microbiome. Tailored dietary or therapeutic interventions targeting the microbiome could advance personalized medicine; however, predicting such interventions requires predictive systems biology methods. Constraint-Based Reconstruction and Analysis (COBRA) is a mechanistic systems biology approach that relies on detailed genome-scale reconstructions of a target organism's metabolism. A resource of genome-scale reconstructions of human microbes, AGORA, and its expansion in size and scope, AGORA2, have been developed through a semi-automated refinement pipeline, DEMETER. A user-friendly analysis pipeline, mgPipe, allows building and interrogating personalized models of microbiome communities from AGORA and AGORA2. Through sample-specific simulations, mgPipe can stratify patients and controls by the distinct metabolic capabilities of their microbiomes, starting from the processed metagenomic sequencing data. Building on this functionality, the protocol provides a comprehensive workflow for the contextualization of metagenomics data through personalized, mechanistic modeling. Comprehensive tutorials for the DEMETER and mgPipe workflows are presented, which will enable both systems biologists and microbiome scientists to contextualize metagenomic data and perform mechanistic simulations of diet-microbiome-host interactions.