Abstract
Introduction: Many missense variants remain categorized as variants of uncertain significance (VUS) due to limited evidence. VUS are at the core of disparities, as underrepresented individuals tend to receive more VUS. To improve variant classification (VC) across race, ethnicity and ancestry (REA) groups, we developed machine learning (ML) models by leveraging genomic datasets. For this study, we evaluated the impact of ML-informed VC for cardiomyopathy, arrhythmia, lipidemia, aortopathy, hereditary hemorrhagic telangiectasia, and pulmonary hypertension genes. Research Question: What is the utility of ML models for cardiovascular disease panel testing VC across REA groups? Methods: From 1/1/2022 to 5/22/2023, ML algorithms that leveraged SpliceAI, gnomAD, AlphaFold and others were validated. ML evidence that met a negative or positive predictive value >80% was incorporated into Sherloc, an ACMG/AMP-based VC framework. At least 1 validated model was available for 225 genes. Analyses of >20,000 patients were performed from random sampling of 20,000 patients with extrapolation. Measurement error was less than 2% variation by bootstrapping. VCs impacted by ML were evaluated, stratified by REA. Results: Out of 113,206 (32% were from an underrepresent group) US-based individuals that underwent panel genetic testing, ~65,900 (58%) had ML evidence applied to at least 1 variant. ML contributed to classifying at least 1 benign/likely benign (B/LB) variant in ~33,200 (29%) and at least 1 pathogenic/likely pathogenic (P/LP) variant in ~2,700 (2.4%) individuals. A higher percentage of Asian (43%), Black (42%), and Hispanic (37%) individuals had an ML-dependent definitive VC (P/LP or B/LB) relative to White (28%) individuals by a one-tailed, two-sample proportion test (p<2.0x10 -58 ). Conclusions: ML modeling demonstrated utility in cardiology diagnostic gene panel testing. Among individuals who had at least 1 variant with ML evidence applied, 54% resulted in definitive classifications. Additionally, ML appears to provide more definitive VCs for underrepresented individuals. ML can assess factors in ways that are agnostic to population ancestry and, when appropriately incorporated into VC, can narrow the VUS gap rates across REA groups.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.