GAMclust: identification of regulated metabolic modules in bulk, single cell and spatial gene expression data.

Abstract

Metabolism operates as a highly interconnected biochemical network, and its regulation emerges from coordinated changes across many reactions and metabolites. The integration of gene expression profiling data with organism-scale metabolic networks has proven to be a valuable tool for understanding cellular metabolic regulation. However, the increasing complexity of profiling technologies and experimental designs requires the development of specialized tools. Here, we present GAMclust, an R package implementing and extending the previously published GAM-clustering pipeline for identifying transcriptionally regulated metabolic modules in complex gene expression datasets. GAMclust supports bulk, single-cell, and spatial gene expression profiling. It includes built-in KEGG and Rhea metabolic networks for human and mouse, with options to expand these networks for the analysis of other species. The package also offers a suite of post-processing and visualization tools, facilitating the exploration and interpretation of results. GAMclust is freely available at https://github.com/alserglab/GAMclust and https://doi.org/10.5281/zenodo.21432552 under the MIT license. Documentation is available at https://alserglab.github.io/GAMclust. Source code for supplementary materials is available at https://github.com/alserglab/GAMclust-paper. Supplementary data are available online.