Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 42 min ago
Peer-reviewedCC BYNature Communications

Why variant effect predictors and multiplexed assays agree and disagree

  • Benjamin Livesey1,2
  • Joseph A. Marsh3,2
  1. 1MRC Human Genetics Unit, Institute of Genetics and Cancer, University of Edinburgh, Edinburgh, UK. [email protected]
  2. 2MRC Human Genetics Unit, Institute of Genetics and Cancer, University of Edinburgh, Edinburgh, UK
  3. 3MRC Human Genetics Unit, Institute of Genetics and Cancer, University of Edinburgh, Edinburgh, UK. [email protected]
Open at sourcePDFDOICite

Abstract

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.

Related

  1. Benchmarking the AlphaGenome Variant Impact score against SpliceAI on post-release ClinVar classifications and experimental splicing assays
    O. A. Cavus, A. Kuskucu
    bioRxivPreprintCC BY-NCFull text1 citedGenetics & genomicsbioRxiv
  2. Variant effect predictors are shaping the ClinVar classifications used to train and evaluate them
    M. Tao, J. A. Marsh
    bioRxivPreprintCC BYFull textGenetics & genomicsbioRxiv
  3. Leakage-Controlled Benchmarking of Variant Effect Predictors and Development of Gradient-Boosting Consensus Model
    M. Havlasek, P. Kabourek, M. Musil +3
    bioRxivPreprintCC BYFull textBioinformatics & computational biologybioRxiv
  4. Combining MAVEs and computational predictors improves variant classification across ancestries in hereditary cancer genes
    Nikita Bora, Mihaly Badonyi, Ankit K. Pathak +4
    Genome MedicinePeer-reviewedCC BY-NC-ND 4.0Full textGenetics & genomicsPMC
  5. Homology-Based Variant-Effect Predictors Break Down on Cytochrome P450 Pharmacogenes
    H. Xu, I. Samori, G. Nayar +1
    bioRxivPreprintCC BYFull textBioinformatics & computational biologybioRxiv

Cite

BibTeXRIS
BibTeX
@article{livesey2026variant,
  title = {{Why variant effect predictors and multiplexed assays agree and disagree}},
  author = {Benjamin Livesey and Joseph A. Marsh},
  journal = {Nature Communications},
  year = {2026},
  month = aug,
  doi = {10.1038/s41467-026-77211-x},
  url = {https://doi.org/10.1038/s41467-026-77211-x},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Why variant effect predictors and multiplexed assays agree and disagree
AU  - Benjamin Livesey
AU  - Joseph A. Marsh
PY  - 2026
DA  - 2026/08/27
JO  - Nature Communications
DO  - 10.1038/s41467-026-77211-x
UR  - https://doi.org/10.1038/s41467-026-77211-x
AB  - 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.
SN  - 2041-1723
ER  -