Disease-specific variant pathogenicity prediction significantly improves variant interpretation in inherited cardiac conditions
Problem
Existing machine learning tools imprecisely discriminate benign from pathogenic rare variants in inherited cardiac conditions.
Why it matters
Accurate variant interpretation is a priority for clinical genome interpretation and patient outcome stratification.
What the AI did
CardioBoost, a disease-specific machine learning classifier, estimates pathogenicity probability for rare missense variants in inherited cardiomyopathies and arrhythmias.
Results
Precision-recall AUC 0.91 (cardiomyopathies) and 0.96 (arrhythmias), outperforming existing tools by 4-24% (metric truncated in abstract).
Impact
CardioBoost demonstrably improves variant discrimination over existing tools for inherited cardiac conditions.
Record
| Venue | Genetics in Medicine |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Xiaolei Zhang, Roddy Walsh, Nicola Whiffin, Rachel J. Buchan, William Midwinter, Alicja Wilk, Risha Govind, Nicholas Li et al. |
| Citations | 63 (OpenAlex, at import time) |
| References | 31 |
| Model or method | CardioBoost |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41436-020-00972-3 |
| Open access | hybrid |
| Keywords | CardioBoost, variant pathogenicity, machine learning, cardiomyopathy, arrhythmia, clinical genomics |
Importance rating
significant — thousands to millions affected
Abstract calls accurate benign/pathogenic discrimination a priority for clinical genome interpretation and patient outcome stratification; benefits patients with inherited cardiac conditions.