Abstract:Proper function of a wastewater treatment plant (WWTP) relies on maintaining a delicate balance between a multitude of competing microorganisms. Gaining a detailed understanding of the complex network of interactions therein is essential to maximising not only current operational efficiencies, but also for the effective design of new treatment technologies. Metagenomics offers an insight into these dynamic systems through the analysis of the microbial DNA sequences present. Unique taxa are deduced through sequence clustering to form operational taxonomic units (OTUs), with per-taxa abundance estimates obtained from corresponding sequence counts. The data in this study comprise weekly OTU counts from an activated sludge (AS) tank of a WWTP along with corresponding measurements of chemical and environmental (CE) covariates. Directly fitting a model to the OTU data is incredibly challenging because of the high dimensionality and sparsity of the observations. The first step is therefore to aggregate the OTUs into twelve microbial communities or "bins" using a seasonal phase-based clustering approach. The mean abundances in the twelve bins are assumed to vary over time according to a multivariate linear regression on the CE covariates. Deviations from the mean are then modelled using a vector autoregressive (VAR) model of order one, which is a linear approximation to the commonly used generalised Lotka-Volterra (gLV) model. Sparsity is assumed in the interactions between microbial communities by carrying out inference in a hierarchical Bayesian framework which uses a shrinkage prior for the autoregressive coefficient matrix of the VAR model. Different shrinkage priors are explored by analysing simulated data sets before selecting the regularised horseshoe prior for the biological application. It is found that ammonia and chemical oxygen demand have a positive relationship with several bins and pH has a positive relationship with one bin. These results are supported by findings in the biological literature. Several negative interactions are also identified. These novel biological findings suggest OTUs in different bins may be competing for resources and that these relationships are complex. Although simpler than a gLV model, the VAR model is still able to offer valuable insight into the microbial dynamics of the WWTP.

Microbiome dynamics analysis using a novel multivariate vector autoregression model with weighted fusion regularization

Weighted fusion regularisation and predicting microbial interactions with vector autoregressive model.

Inference of microbial interactions using multivariate vector auto-regression model

Inference Of Microbial Interactions From Time Series Data Using Vector Autoregression Model

Predicting Microbial Interactions by Using Network-Constrained Regularization Incorporating Covariate Coefficients and Connection Signs

Manifold-constrained Regularization for Variable Selection in Envrionmental Microbiomic Data

A sparse Bayesian hierarchical vector autoregressive model for microbial dynamics in a wastewater treatment plant

MZINBVA: Variational Approximation for Multilevel Zero-Inflated Negative-Binomial Models for Association Analysis in Microbiome Surveys

Variable selection for sparse Dirichlet-multinomial regression with an application to microbiome data analysis

A Bayesian Zero-Inflated Negative Binomial Regression Model for the Integrative Analysis of Microbiome Data

Conditional regression based on a multivariate zero-inflated logistic normal model for microbiome relative abundance data

High-dimensional Linear State Space Models for Dynamic Microbial Interaction Networks

Negative Binomial factor regression with application to microbiome data analysis

Factor Augmented Inverse Regression and its Application to Microbiome Data Analysis

Model-Based Microbiome Data Ordination: A Variational Approximation Approach.

Novel Collaborative Weighted Non-negative Matrix Factorization Improves Prediction of Disease-Associated Human Microbes

ARZIMM: A Novel Analytic Platform for the Inference of Microbial Interactions and Community Stability from Longitudinal Microbiome Study

MKMR: a multi-kernel machine regression model to predict health outcomes using human microbiome data

Bayesian compositional regression with microbiome features via variational inference

A two-part mixed-effects model for analyzing longitudinal microbiome compositional data

Microbiome-based disease prediction with multimodal variational information bottlenecks