StrVCTVRE logo


StrVCTVRE predicts the pathogenicity of deletions and duplications that overlap one or more exons in the human genome. Just as there are methods to predict the pathogenicity of missense variants (such as REVEL, CADD, and many others), StrVCTVRE predicts the pathogenicity of deletions and duplications that affect one or more exons. StrVCTVRE was developed by researchers at UC Berkeley and Harvard Medical School. On this website, you can score a single variant with StrVCTVRE, annotate a VCF or BED file with StrVCTVRE, or download StrVCTVRE to run offline.

StrVCTVRE Summary

How StrVCTVRE works

StrVCTVRE learns to identify pathogenic variants using a random forest trained on data from ClinVar, gnomAD, and a recent great ape sequencing study. StrVCTVRE integrates features that capture gene importance, coding regions, conservation, expression, and exon structure. StrVCTVRE performs accurately across a wide SV size range on independent test sets. We anticipate clinicians and researchers will use StrVCTVRE to prioritize structural variants identified during exome or genome sequencing in rare disease probands where no structural variant is immediately compelling.

StrVCTVRE Features
If you use StrVCTVRE in your published work, please cite:
Sharo AG, Hu Z, Sunyaev SR, and Brenner SE. 2022. StrVCTVRE: A supervised learning method to predict the pathogenicity of human genome structural variants. The American Journal of Human Genetics 109:195-209. doi:10.1016/j.ajhg.2021.12.007