Automated classification of ~7,000 variants shows near perfect concordance with expert panel assessments

Introduction: Gathering the most current and accurate information is critical to variant interpretation. The QIAGEN knowledgebase includes a manually curated database of variant specific publications, and data from public and proprietary databases. This resource is the cornerstone of QIAGEN Clinical Insight (QCI), which facilitates automated variant classification using all 28 ACMG rules, while transparently providing the underlying evidence. Here, we compare the concordance of QCI’s automated variant classification with expert panel assessments across multiple disease indications.

Materials & Methods: Expert panel (ENIGMA) reviewed BRCA1 and BRCA2 variants (n=6154), and variants reviewed by the ClinGen Inherited Cardiomyopathy Expert Panel (n=102) were exported from ClinVar. Resulting VCFs were uploaded into QCI, and concordance of automated classifications was compared with expert panel assessments in ClinVar.

Results: With respect to clinical actionability, automated variant classifications were extremely concordant, reaching 99.6% concordance with ENIGMA assessments of BRCA variants, and 96.1% concordance with cardiomyopathy variants assessed by the ClinGen expert panel. The small number of differences seen in the ENIGMA dataset could be attributed to functional studies used by the automated algorithm not considered by the expert panel. The 3.9% of discrepant classifications in the cardiomyopathy set likely result from differences in clinical case curation.

Conclusions: Through updated content, and continued alignment with professional guidelines, automated variant classification in QCI demonstrates extremely high accuracy across multiple disease contexts. This level of accuracy speaks to the quality of the clinical, functional and population level data curation, as well as the robustness of the underlying ACMG classification algorithm.

Jennifer Poitras

Jennifer received a B.S. in Genetics from the University of Connecticut and went on to receive a Ph.D. in Human Genetics at the Johns Hopkins School of Medicine. During her graduate training, she used Ingenuity Pathway Analysis and was so impressed with the tool that she approached QIAGEN for the opportunity to support their bioinformatics portfolio. Currently, she is the lead genome scientist supporting QIAGEN’s software solutions for analysis and interpretation of variants implicated in hereditary disease.