MDiNE: a model to estimate differential co-occurrence networks in microbiome studies

Abstract

Motivation The human microbiota is the collection of microorganisms colonizing the human body, and plays an integral part in human health. A growing trend in microbiome analysis is to construct a network to estimate the co-occurrence patterns among taxa through precision matrices. Existing methods do not facilitate investigation into how these networks change with respect to covariates. Results We propose a new model called Microbiome Differential Network Estimation (MDiNE) to estimate network changes with respect to a binary covariate. The counts of individual taxa in the samples are modeled through a multinomial distribution whose probabilities depend on a latent Gaussian random variable. A sparse precision matrix over all the latent terms determines the co-occurrence network among taxa. The model fit is obtained and evaluated using Hamiltonian Monte Carlo methods. The performance of our model is evaluated through an extensive simulation study and is shown to outperform existing methods in terms of estimation of network parameters. We also demonstrate an application of the model to estimate changes in the intestinal microbial network topology with respect to Crohn’s disease. Availability and implementation MDiNE is implemented in a freely available R package: https://github.com/kevinmcgregor/mdine. Supplementary information Supplementary data are available at Bioinformatics online.

Network analysis methods for studying microbial communities: A mini review

“…Identification of new biomarkers such as Enterobacteriacea, more abundant in Crohn’s samples and Lachnospiraceae to be less [60] …”

“…Microbiome Differential Network Estimation (MDINE) [60] generates differential networks to show how microbial relationships vary between two conditions based on an estimation of the precision matrix. MDiNE addresses compositionality by utilizing a Dirichlet-multinomial logistic-normal distribution model [61] , [62] .…”

Linking Plant Secondary Metabolites and Plant Microbiomes: A Review

“…For processing the bacterial 16S rRNA gene and fungal ITS amplicons, a collection of software, such as QIIME ( Caporaso et al, 2010 ), UPARSE ( Edgar, 2013 ), VSEARCH ( Rognes et al, 2016 ), PIPITS ( Gweon et al, 2015 ), and USEARCH ( Edgar and Flyvbjerg, 2015 ) have been developed. Similarly, for shotgun microbiome sequencing analyses, several recent articles reported specific computational workflow and bioinformatics resources ( Liu Y. X. et al, 2020 ), including Microbiome Helper ( Comeau et al, 2017 ), HmmUFOtu ( Zheng et al, 2018 ), iMicrobe ( Youens-Clark et al, 2019 ), MMinte ( Mendes-Soares et al, 2016 ), MDiNE ( McGregor et al, 2020 ), MicrobiomeAnalyst ( Dhariwal et al, 2017 ), SIMBA ( Mariano et al, 2016 ), and iMAP ( Buza et al, 2019 ). Several in-depth summaries and comparisons of next-generation amplicon sequencing and analyses approaches were published recently ( Lucaciu et al, 2019 ; Nilsson et al, 2019 ).…”

Editorial: Compositional data analysis and related methods applied to genomics—a first special issue fromNAR Genomics and Bioinformatics

“…Sequencing experiments produce relative count data , and although their nature differs from simple percentages, the application of CoDA has already shown to be advantageous for the analysis of transcriptome and microbiome data. Recent applications include reference-aware analysis of microbial compositions ( 2 ), their dynamics ( 3 ) and phylogenetic scales ( 4 ), reference-aware analysis ( 5 ) and simulation ( 6 ) of RNA-seq data, PCR bias correction ( 7 ), association ( 8 ) and differential network analysis ( 9 ) as well as feature selection ( 10 ) and model fitting ( 11 ). All these techniques apply more generally to positive-valued signal data, implying they could also be used in fields like proteomics ( 12 ) and metabolomics ( 13 ).…”