Nature Communications· 2026Q1
Genomic heterogeneity inflates the performance of variant pathogenicity predictions
- 3citations
- Q1SCImago
- 2026year
Short summary
Performance metrics for genome-wide variant pathogenicity prediction models are severely inflated due to the mixing of variant categories with different baseline prevalences of pathogenic variants.
AI-generated from the title and abstract; the full text is not read.
Key points
- Genome-wide variant pathogenicity prediction models show inflated performance metrics.
- Inflation is caused by mixing variant categories with different baseline pathogenic prevalences.
- A new benchmark stratified by variant type is proposed for accurate model evaluation.
- Reliable predictions are still out of reach for certain genomic contexts.
AI-generated from the title and abstract; the full text is not read.
Abstract
Abstract Recent studies have reported unprecedented accuracy predicting pathogenic variants across the genome, including in noncoding regions, using large deep-learning models trained on vast genomic data. To verify and understand the basis of these reports, we systematically evaluated the prediction accuracy of frontier models over single-nucleotide variants, focusing on general models capable of scoring multiple variant types across coding or noncoding regions. Our evaluations reveal that the performance of genome-wide predictions is severely inflated by mixing variant categories with different baseline prevalences of pathogenic variants. We constructed a benchmark stratified by variant type to identify the best-performing models for each genomic context, determine where reliable predictions remain out of reach, and establish rigorous standards to guide future evaluations.
The authors' abstract, as published at the source. Nature Communications, 2026 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
Field: Genetics (Biochemistry, Genetics and Molecular Biology)
GeneticsBiochemistry, Genetics and Molecular Biology