ggClusterNet 2: An R package for microbial co-occurrence networks and associated indicator correlation patterns

Abstract

Since its initial release in 2022, ggClusterNet has become a vital tool for microbiome research, enabling microbial co-occurrence network analysis and visualization in over 300 studies. To address emerging challenges, including multi-factor experimental designs, multi-treatment conditions, and multi-omics data, we present a comprehensive upgrade with four key components: (1) A microbial co-occurrence network pipeline integrating network computation (Pearson/Spearman/SparCC correlations), visualization, topological characterization of network and node properties, multi-network comparison with statistical testing, network stability (robustness) analysis, and module identification and analysis; (2) Network mining functions for multi-factor, multi-treatment, and spatiotemporal-scale analysis, including Facet.Network() and module.compare.m.ts(); (3) Transkingdom network construction using microbiota, multi-omics, and other relevant data, with diverse visualization layouts such as MatCorPlot2() and cor_link3(); and (4) Transkingdom and multi-omics network analysis, including corBionetwork.st() and visualization algorithms tailored for complex network exploration, including model_maptree2(), model_Gephi.3(), and cir.squ(). The updates in ggClusterNet 2 enable researchers to explore complex network interactions, offering a robust, efficient, user-friendly, reproducible, and visually versatile tool for microbial co-occurrence networks and indicator correlation patterns. The ggClusterNet 2R package is open-source and available on GitHub (https://github.com/taowenmicro/ggClusterNet).