Description

A multiplexed assay of variant effect (MAVE), such as deep mutational scanning or saturation genome editing, measures the functional effect of thousands of variants of one gene or element in a single experiment. MaveDB is the public repository for these data sets.

This track collection shows the human MaveDB score sets on hg38. It has three tracks:

• MaveDB Heatmaps: one heatmap per score set, with a column for each measured codon (or base) and a row for each substitution.
• MaveDB Variants: one item per measured variant per score set, searchable by variant name (for example TP53:R175H), ClinGen allele ID or MaveDB variant ID.
• MaveDB Alignments: where the sequence each experiment tested lies on the genome.

The data are refreshed from MaveDB automatically when MaveDB adds or changes score sets; the version line on each track's page gives the date of the copy shown.

Only variants that could be placed on the genome are shown. Most are placed by MaveDB's own mapping. Where MaveDB has no mapping, or where its mapping disagrees with the amino acids the experiment reports, the variant is placed by UCSC from the experiment's own description and checked against the reference protein and genome; the track details say which applies to each variant, and the MaveDB Variants page describes how. Score sets measured on designed or heavily modified proteins that cannot be matched to a reference protein, and score sets of which only a small fraction of variants could be placed, are not shown. Variants that change several sites at once, and variants with no score, are not shown.

Scores are as submitted to MaveDB. Each score set has its own scale and direction: in some a high score means a normal protein, in others an abnormal one. Compare variants within a score set, not between score sets. For clinically calibrated interpretations of a subset of these score sets, see the MaveMD track collection.

Data Access

The pages of the individual tracks describe how to download and query their data.

Credits

Thanks to the MaveDB team, in particular Alan Rubin and Ben Capodanno, and to the many labs who deposited their data in MaveDB. The first version of these tracks was built by Jonathan Casper at UCSC.

References

Arbesfeld JA, Da EY, Stevenson JS, Kuzma K, Paul A, Farris T, Capodanno BJ, Grindstaff SB, Riehle K, Saraiva-Agostinho N et al. Mapping MAVE data for use in human genomics applications. Genome Biol. 2025 Jun 25;26(1):179. PMID: 40563119; PMC: PMC12188674

Rubin AF, Stone J, Bianchi AH, Capodanno BJ, Da EY, Dias M, Esposito D, Frazer J, Fu Y, Grindstaff SB et al. MaveDB 2024: a curated community database with over seven million variant effects from multiplexed functional assays. Genome Biol. 2025 Jan 21;26(1):13. PMID: 39838450; PMC: PMC11753097