Longitudinal Prediction of the Infant Gut Microbiome with Dynamic Bayesian Networks.

Michael J Mcgeachie,George M Weinstock,Diane R Gold,Travis Gibson,Scott T Weiss,Yang-Yu Liu,Joanne E Sordillo,Augusto Litonjua

doi:10.1038/srep20359

Michael J Mcgeachie, George M Weinstock + Show 6 more

Open Access

https://doi.org/10.1038/srep20359

Copy DOI

Abstract

Sequencing of the 16S rRNA gene allows comprehensive assessment of bacterial community composition from human body sites. Previously published and publicly accessible data on 58 preterm infants in the Neonatal Intensive Care Unit who underwent frequent stool collection was used. We constructed Dynamic Bayesian Networks from the data and analyzed predictive performance and network characteristics. We constructed a DBN model of the infant gut microbial ecosystem, which explicitly captured specific relationships and general trends in the data: increasing amounts of Clostridia, residual amounts of Bacilli, and increasing amounts of Gammaproteobacteria that then give way to Clostridia. Prediction performance of DBNs with fewer edges were overall more accurate, although less so on harder-to-predict subjects (p = 0.045). DBNs provided quantitative likelihood estimates for rare abruptions events. Iterative prediction was less accurate (p < 0.001), but showed remarkable insensitivity to initial conditions and predicted convergence to a mix of Clostridia, Gammaproteobacteria, and Bacilli. DBNs were able to identify important relationships between microbiome taxa and predict future changes in microbiome composition from measured or synthetic initial conditions. DBNs also provided likelihood estimates for sudden, dramatic shifts in microbiome composition, which may be useful in guiding further analysis of those samples.

Highlights

The microbiota living in the human gut performs a number of vital functions for homeostasis, including the harvest of essential nutrients[1,2], synthesis of vitamins[3], metabolism of xenobiotics[4], and the development and maintenance of the immune system[5,6]
With 922 microbiome assessments, this is an attractive cohort for Dynamic Bayesian Networks (DBN) analysis, of which other authors have suggested that 1000 samples is a good baseline for reconstructing an adequately accurate DBN25 La Rosa et al used this data to demonstrate the succession of bacterial colonization, and showed the importance of post-conceptional age to the developing infant gut microbiome
With a DBN model, we were able to: (1) identify how the three dominant bacterial classes influence one another over time; (2) account for the role of relatively rare taxa in influencing these dominant groups; (3) determine the importance of initial conditions for iterative prediction of the post-natal microbiome trajectory; and (4) identify samples that depart from the expected trajectory of infant gut microbiome development in the first month of life

Summary

Introduction

The microbiota living in the human gut performs a number of vital functions for homeostasis, including the harvest of essential nutrients[1,2], synthesis of vitamins[3], metabolism of xenobiotics[4], and the development and maintenance of the immune system[5,6]. GLV has been used in a study of the murine gut microbiome to generate a network of interactions between bacterial taxa[18]; this network included almost all possible edges and was not used for prediction. In another microbiome study, a continuous GLV model was assumed and coefficients related to the individual microbe growth rates, the strengths of the microbe-microbe interactions, and susceptibility to antibiotics were learned using linear regression with regularization[17]. One such method used Bayesian statistics to help inform dynamic models of a single independent bacteria taxon’s change in response to antibiotics[21]

Methods

Results

Conclusion