Mounting studies in humans, animals, and cells have shown that microbiome and metabolome are closely linked to human health and their compositional and functional abnormalities can lead to a variety of diseases. The selection and jointly usage of multiple methods on cross-omics correlation analysis is a challenging work. It is also challenging to identify key metabolites and microbes from huge amounts of results under the instruction of metadata (phenotypes and/or confounders).

The metabolome-microbiome-metadata correlation analysis platform (3MCor) is a web server for integrative correlation analysis of metabolome and microbiome under the instruction of metadata (phenotypes and confounders). It is independent to measurement platforms and sample types. Microbial data sets (OTU/ASV/taxa/function) from amplicon sequencing (16S, 18S, ITS) or shotgun metagenomic sequencing and metabolomic data sets (qualitative or quantitative) from mass spectrometry or NMR are all acceptable. Metadata including phenotypes (e.g. case-control, disease stages, etc.) and/or confounders (e.g. age, sex, BMI, diet, etc.) are optional inputs. Many univariate methods including generalized linear regression (GLM), (partial) Pearson, (partial) Spearman, maximal information coefficient (MIC), generalized correlation analysis for metabolome and microbiome (GRaMM), Sparse Correlations for Compositional data (SparCC), and Correlation inference for Compositional data through Lasso (CCLasso) are available for the intra-correlation pipeline. Three inter-correlation pipelines, for global (matrix-matrix), hierarchical (cluster-cluster), and pairwise (metabolite-microbial taxa/function) correlation analysis, are provided. Comprehensive numerical results (.csv) and rich figures (.pdf) will be generated in minutes (usually 1~3 mins).

Cite: Tao Sun, Mengci Li, Xiangtian Yu, Dandan Liang, Guoxiang Xie, Chao Sang, Wei Jia, Tianlu Chen, 3MCor: an integrative web server for metabolome–microbiome-metadata correlation analysis, Bioinformatics, 2021;, btab818, https://doi.org/10.1093/bioinformatics/btab818