Capturing the dynamics of microbial interactions through individual-specific networks

Abstract

Longitudinal analysis of multivariate individual-specific microbiome profiles over time or across conditions remains dauntin. Most statistical tools and methods that are available to study microbiomes are based on cross-sectional data. Over the past few years, several attempts have been made to model the dynamics of bacterial species over time or across conditions. However, the field needs novel views on handling microbial interactions in temporal analyses. This study proposes a novel data analysis framework, MNDA, that combines representation learning and individual-specific microbial co-occurrence networks to uncover taxon neighborhood dynamics. As a use case, we consider a cohort of newborns with microbiomes available at 6 and 9 months after birth, and extraneous data available on the mode of delivery and diet changes between the considered time points. Our results show that prediction models for these extraneous outcomes based on an MNDA measure of local neighborhood dynamics for each taxon outperform traditional prediction models solely based on individual-specific microbial abundances. Furthermore, our results show that unsupervised similarity analysis of newborns in the study, again using the notion of a taxon's dynamic neighborhood derived from time-matched individual-specific microbial networks, can reveal different subpopulations of individuals, compared to standard microbiome-based clustering, with potential relevance to clinical practice. This study highlights the complementarity of microbial interactions and abundances in downstream analyses and opens new avenues to personalized prediction or stratified medicine with temporal microbiome data.

netMUG: a novel network-guided multi-view clustering workflow for dissecting genetic and facial heterogeneity

“…The appropriateness of a measure of association between features is context-dependent. For instance, a microbial co-occurrence network has been utilized for building ISNs with microbiome data ( Yousefi et al, 2023 ).…”

PLEX.I: a tool to discover features in multiplex networks that reflect clinical variation

“…Given a multiplex network with n nodes, PLEX.I first maps the nodes of each layer to a common embedding space of dimension (a user-defined parameter). To obtain a simultaneous embedding of all network layers, we designed an encoder-decoder neural network (EDNN) with -dimensional inputs and outputs, and a -dimensional bottleneck layer ( Ietswaart et al, 2021 ; Yousefi et al, 2023 ). For a discussion of the suitability of this approach, for our application, over other network representation learning methods, we refer to Hamilton et al (2017) , Yousefi et al (2023) .…”

“…In Scenario II, we consider a set of two-layer networks, each for one individual; and, the aim is to detect nodes whose neighborhood variation, from one layer to the other, is associated with a particular phenotype across individuals. We have shown an application of both scenarios on human gut microbiomes data in our previous study ( Yousefi et al, 2023 ).…”

Disentangling the effects of terroir, season, and vintage on the grapevine fungal pathobiome

“…The application of network-based approaches, particularly unweighted and weighted metric analysis, provides an avenue to discern the structure and complexity of these microbial communities. By quantifying the interactions among microbial taxa and assessing their relative importance within the network, this approach helps to elucidate the intricate interplay between microbial taxa and their environment ( Herren and McMahon, 2017 ; Banerjee et al, 2018 ; Yousefi et al, 2023 ). Unweighted network analysis, based on presence/absence of taxa, provides insight into community structures and interactions that may be driven by environmental filters or competitive exclusion.…”

“…Unweighted network analysis, based on presence/absence of taxa, provides insight into community structures and interactions that may be driven by environmental filters or competitive exclusion. In contrast, weighted networks, accounting for the abundance of different taxa, enable detection of more subtle and potentially important interactions that are often overlooked in unweighted analysis, including those influenced by the relative abundance of taxa ( Barberán et al, 2012 ; Yousefi et al, 2023 ). This methodology allows for a deeper understanding of community structures and interactions that may be driven by environmental filters or competitive exclusion ( Barberán et al, 2012 ; Banerjee et al, 2018 ; Yousefi et al, 2023 ).…”