{"database":"biostudies-literature","file_versions":[],"scores":null,"additional":{"submitter":["Martin C"],"funding":["U.S. Department of Health & Human Services | NIH | National Institute of General Medical Sciences","U.S. Department of Health & Human Services | NIH | National Institute of General Medical Sciences (NIGMS)","NIGMS NIH HHS"],"pagination":["11123"],"full_dataset_link":["https://www.ebi.ac.uk/biostudies/studies/S-EPMC12705698"],"repository":["biostudies-literature"],"omics_type":["Unknown"],"volume":["16(1)"],"pubmed_abstract":["Exponential increases in microbial and viral genomic data demand transformational advances in scalable, generalizable frameworks for their interpretation. Standard homology-based functional analyses are hindered by the rapid divergence of microbial and especially viral genomes and proteins that significantly decreases the volume of usable data. Here, we present Protein Set Transformer (PST), a protein-based genome language model that models genomes as sets of proteins without considering sparsely available functional labels. Trained on >100k viruses, PST outperforms other homology- and language model-based approaches for relating viral genomes based on shared protein content. Further, PST demonstrates protein structural and functional awareness by clustering capsid-fold-containing proteins"],"journal":["Nature communications"],"pubmed_title":["Protein Set Transformer: a protein-based genome language model to power high-diversity viromics."],"pmcid":["PMC12705698"],"funding_grant_id":["R35 GM143024","R35GM143024"],"pubmed_authors":["Anantharaman K","Gitter A","Martin C"],"additional_accession":[]},"is_claimable":false,"name":"Protein Set Transformer: a protein-based genome language model to power high-diversity viromics.","description":"Exponential increases in microbial and viral genomic data demand transformational advances in scalable, generalizable frameworks for their interpretation. Standard homology-based functional analyses are hindered by the rapid divergence of microbial and especially viral genomes and proteins that significantly decreases the volume of usable data. Here, we present Protein Set Transformer (PST), a protein-based genome language model that models genomes as sets of proteins without considering sparsely available functional labels. Trained on >100k viruses, PST outperforms other homology- and language model-based approaches for relating viral genomes based on shared protein content. Further, PST demonstrates protein structural and functional awareness by clustering capsid-fold-containing proteins","dates":{"release":"2025-01-01T00:00:00Z","publication":"2025 Nov","modification":"2026-06-06T03:59:03.985Z","creation":"2026-05-25T03:11:54.259Z"},"accession":"S-EPMC12705698","cross_references":{"pubmed":["41276529"],"doi":["10.1038/s41467-025-66049-4"]}}