Skip to main content
eScholarship
Open Access Publications from the University of California

Learning representations of microbe-metabolite interactions.

  • Author(s): Morton, James T
  • Aksenov, Alexander A
  • Nothias, Louis Felix
  • Foulds, James R
  • Quinn, Robert A
  • Badri, Michelle H
  • Swenson, Tami L
  • Van Goethem, Marc W
  • Northen, Trent R
  • Vazquez-Baeza, Yoshiki
  • Wang, Mingxun
  • Bokulich, Nicholas A
  • Watters, Aaron
  • Song, Se Jin
  • Bonneau, Richard
  • Dorrestein, Pieter C
  • Knight, Rob
  • et al.
Abstract

Integrating multiomics datasets is critical for microbiome research; however, inferring interactions across omics datasets has multiple statistical challenges. We solve this problem by using neural networks (https://github.com/biocore/mmvec) to estimate the conditional probability that each molecule is present given the presence of a specific microorganism. We show with known environmental (desert soil biocrust wetting) and clinical (cystic fibrosis lung) examples, our ability to recover microbe-metabolite relationships, and demonstrate how the method can discover relationships between microbially produced metabolites and inflammatory bowel disease.

Main Content
Current View