NIH grant: Predicting, designing, and controlling microbiome metabolism (1R35GM164154-01)

retained text

NIH grant 1R35GM164154-01: Predicting, designing, and controlling microbiome metabolism
Institute: {'code': 'GM', 'abbreviation': 'NIGMS', 'name': 'National Institute of General Medical Sciences', 'admin_org_id': '12500024', 'admin_funding_url': ''}. PI: Seppe  Kuehn. Award: $149650.
Start: 2026-07-16T00:00:00.
Project Summary/Abstract
Microbial communities drive metabolic processes across environments that are essential to human health,
from the gut and skin to soils. These communities influence immune regulation, resistance to pathogens,
nutrient cycling, and pollutant degradation. Their metabolism affects key health outcomes, including
inflammation, nutrient absorption, and exposure to toxins. The ability to predict, design, or control microbiome
metabolism is limited due to ecosystem complexity.
A major bottleneck is the limited understanding of microbial traits—the metabolic capabilities and physiological
constraints that govern behavior. While genome and protein data are abundant, trait-level knowledge is sparse
and biased toward model organisms. This program addresses that gap by combining microbial physiology with
quantitative statistical and modeling approaches to untangle microbiome complexity and build generalizable
insights into collective metabolism.
Project 1 develops neural network-powered methods to learn metabolic guilds, groups of microbes with
shared functional roles, from sequencing and metabolite data. These guilds serve as interpretable variables in
models linking community composition to metabolism. The approach builds on prior work in denitrification and
extends to complex pathways including denitrification cascades and anaerobic polysaccharide degradation.
The project tests whether gene, transcript, or enzyme abundances outperform species-level data in predicting
metabolism.
Project 2 uses community-function landscapes to design synthetic consortia for pathogen inhibition and
pollutant degradation. It examines how environmental stress promotes cooperation and epistasis in
community-function landscapes, and whether functional redundancy enables consortia to generalize across
environments, a key step toward microbiome-based therapeutics.
Project 3 seeks to uncover principles of resource-based host–microbiome interactions. It investigates host
control of microbiomes in external niches, such as the skin, rhizosphere, or phycosphere, where high dispersal
and weak physical barriers limit selective filtering. Using Chlamydomonas reinhardtii as a model, the project
examines whether bacterial protein degradation supplies nitrogen to stressed algae. It tests whether hosts
passively recruit ammonium-producing bacteria via carbon exudates or actively bias community assembly, and
how cooperative nutrient exchange persists despite selfish microbial strategies.
By integrating physiology, machine learning, modeling, and high-throughput experiments, this program
establishes predictive, mechanistic frameworks for understanding and engineering microbial community
metabolism in health-relevant contexts.

document fetched 2026-07-26 08:23