Phylogenetically informed Bayesian truncated copula graphical models for microbial association networks

Microorganisms play critical roles in host health. The advancement of high-throughput sequencing technology provides opportunities for a deeper understanding microbial interactions. However, due to the technological limitations 16S ribosomal RNA sequencing, microbiome data are zero-inflated, and quantitative comparison abundances cannot be made across subjects. By leveraging recent profiling technique that quantifies counts, we propose novel Bayesian graphical model incorporates microorganisms’ evolutionary history through phylogenetic tree prior explicitly accounts zero inflation using truncated Gaussian copula. Our simulation study reveals information substantially improves network estimation accuracy. We apply proposed gut 106 healthy subjects identify three distinct communities not found by existing models. further find these discriminated based on ability utilize oxygen as an energy source.