Quantifying Taxonomic Uncertainty in Metabolic Modeling using Bayesian Inference Method
Lacy, Jacob
Lacy, Jacob
Citations
Altmetric:
License
License
Abstract
Chronic inflammatory diseases, such as Ulcerative Colitis (UC), are multifaceted issues with dysbiosis of the gut microbiome often being associated. Genome-Scale Metabolic Models (GEMs) can provide insights into metabolic outputs of the microbiome; however, a major challenge in their applications to microbiome modeling is taxonomic uncertainty in microbiome data. Which is typically mapped to the genus or species level, while GEMs are strain (genome) specific. Here we have developed a Bayesian inference method to quantify uncertainty in GEMs simulations at different taxonomic levels. We used 7301 Assembly of Gut Organisms Reconstruction and Analysis version 2 (AGORA 2) models and simulated metabolism with parsimonious Flux Balance Analysis (pFBA). We established a global metabolic prior at the domain taxonomic rank that is recursively updated down the taxonomic ladder using a Student�s t-distribution likelihood. We applied this method to individual organisms and sequencing data from stool samples collected in a UC fecal microbiota transplant clinical trial. Our framework showed that metabolic uncertainty is not uniformly distributed across the microbiome. For UC associated metabolites such as butyrate and acetate, some microbes produce high precision distributions while others produce broad multimodal metabolic output even at the genus level. The Bayesian approach provides uncertainty aware prediction of metabolic outputs that can be used for reliable inference of metabolic differences between diseased and healthy states.
Description
Date
1/1/2026
Student Status
Graduate Student
Journal Title
Journal ISSN
Volume Title
Type of presentation
Poster
Collections
Research Projects
Organizational Units
Journal Issue
Citation
DOI
Advisor(s)
Department
Program/Major
Biomedical Engineering
College/School
College of Engineering and Mathematical Sciences
Organization
Research Category
Engineering