MetaDome 2027: a comprehensively updated resource for aggregating missense variant evidence across homologous human protein domains

MetaDome maps genetic variation onto homologous human protein domains, so that variants scattered across different genes can be compared at evolutionarily equivalent positions. This record holds the complete MetaDome v2.0 data release for the GRCh37 and GRCh38 human reference assemblies. Two files serve most use cases. The final dataset (..._final-dataset-sw10.tsv.gz) is the primary product. It gives one row per codon, carrying the tolerance landscape score, the Pfam domain and consensus position, and the pathogenic ClinVar variation found at evolutionarily equivalent positions across the proteome. It is the file to use for variant interpretation, for meta-domain aggregation, or to work with MetaDome without the web platform. The prebuild mapping database (..._prebuild-mapping-database.csv.gz) is the relational mapping beneath it: a per-base correspondence between genomic coordinates, protein-coding transcripts, UniProtKB/Swiss-Prot protein sequences and Pfam protein domains across the human proteome. It contains only mappings where the GENCODE translation and the Swiss-Prot sequence are 100% identical, which is what makes the domain alignment dependable. It loads the MetaDome database and serves as a coordinate-conversion resource in its own right.The remaining files are secondary. Three genome-browser tracks per assembly are derived from the final dataset and contain nothing it does not already hold. Two compressed archives per assembly hold the intermediate prebuild output, included so the release can be reproduced without repeating a multi-day build. Every file derives from a single build per assembly, and each filename states the source versions it was built from.