NACG benchmarking of AI-based variant interpretation solutions

NACG aims to assess the effectiveness of AI-based solutions in prioritizing genetic variants using real-world patient data within clinical settings.

The transition from WES to WGS-based genetic testing for patients suspected of a Mendelian condition is nearly complete in the Nordics. While WGS allows for comprehensive analysis of coding and non-coding regions, the interpretation of genetic variants in relation to a patient's phenotype remains a significant bottleneck. Clinicians must review an average of 100 variants per case, a process that can take over 50 hours per patient. Due to time constraints, only a fraction of variants is typically analyzed, undermining the potential of WGS applications. AI offers a compelling solution to shorten the analysis time and increase diagnostic yield by synergistically merging predictive methodologies with the ever-expanding understanding of genetic diseases. Although market-ready AI-based solutions exist, their adoption in clinical settings remains limited, largely due to a lack of trust from clinicians and hospital managers. This study aims to provide evidence of the effectiveness of AI-based solutions for variant interpretation in clinical genomics to interested stakeholders in Nordic hospitals.

Contact persons: Oleg Agafonov and Ksenia Lavrichenko

Contact person

Placeholder image for Oleg Agafonov

Oleg Agafonov

GRD Healthcare, DNV

Placeholder image for Ksenia Lavrichenko

Ksenia Lavrichenko

Department of Medical Genetics, OUS