Human Proteome Exhaustive Missense Variant Effect Prediction Resource

This dataset provides a proteome-wide exhaustive missense variant effect prediction resource for the human reference proteome (UniProt Proteome ID: UP000005640). The resource comprises predictions for all 19 possible amino acid substitutions at every protein position, covering 20,521 human proteins and a total of 564,287,628 missense variants. Each protein is distributed as an individual TSV file containing the maximum available set of variant effect predictors for that protein. Predictions include scores from the complete QAFI family together with external variant effect predictors integrated into the framework, including AlphaMissense, popEVE, EVE, CPT-1, REVEL, and BayesDel. Coverage varies across proteins depending on the availability of AlphaFold structural models, multiple sequence alignments, and other features required by specific predictors. The computational pipeline used to generate this resource is publicly available at the accompanying GitHub repository. This dataset accompanies the QAFI methodology. The authors acknowledge support from the Ministerio de Ciencia, Innovación y Universidades de España through grant PID2022-142753OB-I00, Ministerio de Ciencia, Innovación y Universidades de España, European Regional Development Fund (FEDER).