I am continuing to watch sessions from the KBase Microbial Community Modeling Workshop last year. Chris Henry from Argonne National Laboratory, presented on “Community Modeling Simulation Tools.” Henry and team work with the GROW project. Henry applied ModelSEED2 to build models of GROW MAGs (Genome Resolved Open Watersheds Metagenome Assembled Genomes). The MAGs are incomplete and have varying levels of contamination. Henry explained that with MAGs there is more gap filling. Dereplicated MAGs from samples will eventually help with model predictions. The simulation approach they are building will use probabilistic models and also integrate energy strategies from every MAG in the community, annotations, and expression data if available. The workflow for the GROW project starts with annotation of MAGs, merging of all annotations in clade, model reconstruction and gapfilling, and community reconstruction to eventually obtain community flux profiles. The results from running GROW project community model of 57 samples indicated clustering of members. Henry shared the “GROW 6 Clade Community Model Analysis” narrative and described the apps used.
