A global genomic and structural resource for the human skin microbiome

Abstract The human skin microbiome plays a critical role in host physiology and immunity, yet its genomic and functional diversity remains under-characterized. Here, we present the Global Skin Metagenomic Data (GSMD) resource, an integrated catalog of 12,409 newly sequenced and 4,396 publicly available skin metagenomes, along with 3,077 isolate genomes. From this dataset, we reconstructed 452,397 metagenome-assembled genomes (MAGs) and identified 11,891 prokaryotic, 165 fungal, and 87,770 viral species-level genome bins (SGBs). This resource introduces 42,758 previously uncharacterized taxa, substantially expanding the known phylogenetic diversity and metabolic repertoire of skin-associated microbial communities. In addition, we cataloged 1.71 billion protein sequences, identifying 79,434 novel sequence clusters lacking homology to existing functional databases. By generating high-confidence three-dimensional structural models for 1,018 of these uncharacterized proteins, we identified unique structural folds that suggest novel biochemical functions. The GSMD resource establishes a high-resolution genomic and structural foundation that facilitates the systematic exploration and standardized interpretation of the human skin microbiome.