Why variant effect predictors and multiplexed assays agree and disagree

Multiplexed assays of variant effect (MAVEs) and computational variant effect predictors (VEPs) are two key tools that provide evidence for the interpretation of genetic variants. While their outputs are often concordant, there are also many differences. Here, we analyse missense MAVE data from 40 different human proteins, comparing them to state-of-the-art VEPs in order to quantify and explain their points of agreement and disagreement. We find that discordance is not random but reflects fundamental differences in how each method infers variant effects. VEPs, which rely heavily on sequence conservation and basic structural features, tend to predict buried and bulky hydrophobic residues as more damaging, while underpredicting impact in disordered regions and at charged surface residues. MAVEs, by contrast, capture context-specific mechanisms more accurately, but can miss damaging variants when the assay fails to reflect disease biology, or be subject to high levels of experimental noise. By comparing both global patterns and specific clinically relevant variants, we show how protein features, assay design, and variant type shape prediction discordance. Our findings provide a framework for interpreting when and why MAVEs and VEPs diverge and point toward strategies for improving variant interpretation through integrated, mechanism-aware approaches.