MaveMD: a functional data resource for genomic medicine

Abstract

Variant interpretation remains one of the most significant challenges in clinical genetics. Variants of uncertain significance (VUS) undermine precision medicine implementation because they have an unknown relationship to disease and cannot be used for clinical decision-making. While evidence from multiplexed assays of variant effect (MAVEs) and other functional assays can help classify variants, major barriers prevent routine use in clinical variant classification, including fragmentation across multiple repositories, insufficient data standards, and the need to calibrate assays clinically. Here we address these challenges by presenting a new interface for the MaveDB database called MaveMD (MAVEs for MeDicine) that integrates with external resources such as ClinVar and the ClinGen Allele Registry, displays clinical evidence calibrations, provides intuitive visualizations, and exports structured evidence compatible with ACMG/AMP variant classification guidelines. MaveMD implements automatic mapping of MaveDB datasets to the human reference genome, dramatically simplifying the clinical translation of new MAVE data. We also defined a new metadata model after curating 438,318 variant effect measurements from 74 MAVE datasets spanning 32 disease-associated genes, and created an interface aimed at enabling effective clinical decision-making. Thus, MaveMD makes MAVE data accessible and easily usable for variant classification, and will scale seamlessly with future data generation efforts to empower the use of MAVE evidence in clinical practice.

The authors' abstract, as published at the source. Genome Medicine, 2026 · DOI ↗