MNetClass: a control-free microbial network clustering framework for identifying central subcommunities across ecological niches.

Investigating microbiome subnetworks and identifying central microbes in specific ecological niches is a critical issue in human microbiome studies. Traditional methods typically require control samples, limiting the ability to study microbiomes at distinct body sites without matched controls. Moreover, some clustering methods are not well-suited for microbial data and fail to identify central subcommunities across ecological niches after clustering. In this study, we present MNetClass, a novel microbial network clustering analysis framework. It utilizes a random walk algorithm and a rank-sum ratio-entropy weight evaluation model to classify key subnetworks and identify central microbes at any body site, without the need for control samples. We demonstrate its capabilities on both simulated and real microbiome data sets. Simulation results indicate that MNetClass outperforms current unsupervised microbial clustering methods. In applied case studies, the analysis of microbiome data from five distinct oral sites revealed site-specific microbial communities. Furthermore, MNetClass demonstrated superior predictive performance on cross-cohort Autism Spectrum Disorder data and identified age-related microbial communities across different oral sites, underscoring its broad applicability in microbiome research.IMPORTANCEMNetClass provides a valuable tool for microbiome network analysis, enabling the identification of key microbial subcommunities across diverse ecological niches. Implemented as an R package (https://github.com/YihuaWWW/MNetClass), it offers broad accessibility for researchers. Here, we systematically benchmarked MNetClass against existing microbial clustering methods on synthetic data using various performance metrics, demonstrating its superior efficacy. Notably, MNetClass operates without the need for control groups and effectively identifies central microbes, highlighting its potential as a robust framework for advancing microbiome research.