Abstract
Microbial community metabolomics, particularly in the human gut, are beginning to provide a new route to identify functions and ecology disrupted in disease. However, these data can be costly and difficult to obtain at scale, while amplicon or shotgun metagenomic sequencing data are readily available for populations of many thousands. Here, we describe a computational approach to predict potentially unobserved metabolites in new microbial communities, given a model trained on paired metabolomes and metagenomes from the environment of interest. Focusing on two independent human gut microbiome datasets, we demonstrate that our framework successfully recovers community metabolic trends for more than 50% of associated metabolites. Similar accuracy is maintained using amplicon profiles of coral-associated, murine gut, and human vaginal microbiomes. We also provide an expected performance score to guide application of the model in new samples. Our results thus demonstrate that this ‘predictive metabolomic’ approach can aid in experimental design and provide useful insights into the thousands of community profiles for which only metagenomes are currently available.
Highlights
Microbial community metabolomics, in the human gut, are beginning to provide a new route to identify functions and ecology disrupted in disease
We have developed MelonnPan as a computational method to predict metabolite features from amplicon or metagenomic sequencing data by incorporating biological knowledge in the form of either taxonomic or functional profiles
MelonnPan represents a newly developed method to infer approximate metabolite feature abundances associated with microbial communities, and its validation and applications show that the information contained in microbiome taxonomic and functional profiles is sufficiently correlated with metabolomic content to infer actionable predictions of microbial community biochemical environments
Summary
In the human gut, are beginning to provide a new route to identify functions and ecology disrupted in disease. Identifying such associations purely based on enzymatic roles is greatly limited by the currently unsaturated repertoire of gene–metabolite reactions, as well as by the relative (rather than absolute) abundance measures provided both by typical sequencing and metabolomic technologies Despite these limitations, approaches that predict metabolite features associated with gut microbial profiles can serve as a hypothesis generator that can facilitate population-scale discovery of novel associations (e.g. in large metagenomic data collections) and lead to new sets of testable hypotheses, serving as a complementary adjunct to experimental validation studies (e.g. as has been the case for predictive functional profiling from amplicon data[20]). The implementation of MelonnPan, associated documentation, and example datasets are made freely available in the MelonnPan software package at http:// huttenhower.sph.harvard.edu/melonnpan
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have