Poster: High-Throughput Genetic Variant Classification for Inherited Cancer Gene Panels Through An AI Inference Engine
Franciso De La Vega, D.Sc.
Chief Scientific Officer, Fabric Genomics
Sahar Nohzadeh-Malakshah, Ph.D.
Senior Bioinformatics Scientist, Fabric Genomics
Friday, November 8, 2019
Baltimore Convention Center Exhibit Hall
A growing number of labs are implementing cancer risk screening tests that sequence panels of cancer genes ranging from a few to over a hundred genes. This growth is driven by both the reduction of sequencing costs and the availability of reimbursement pathways for such tests. However, an important part of the cost involves the assessment of variant pathogenicity by trained clinical geneticists. The ACMG and AMP developed evidence-based guidelines to standardize variant assessment defining several criteria for supporting evidence of pathogenicity, which are then combined to classify a variant as either pathogenic (P), likely-pathogenic (LP), benign (B), likely-benign (LB), or uncertain significance (VUS). Although widely adopted in clinical interpretation of variants this process has remained largely manual and time-consuming.