Platform to support clinical variant interpretation through probabilistic assessment of functional evidence - PROJECT SUMMARY Advances in DNA sequencing technology have revolutionized the diagnosis and treatment of disease, including cancer, cardiovascular disease, and neurodegeneration. Accurate clinical interpretation of genetic variants can enable early detection, inform precision treatment plans, and improve patient outcomes. However, the majority of unique variants uncovered by clinical genetic tests have insufficient evidence for confident classification and are therefore categorized as “variants of uncertain significance” (VUS). The uncertainty introduced by VUS impedes diagnosis, delays care, and increases unnecessary medical treatment. Constantiam Biosciences has developed Varify, an innovative platform that addresses the VUS problem at scale by leveraging the recent advent and rapid proliferation of Multiplexed Assays of Variant Effect, or MAVEs, for variant classification. MAVEs generate functional data for thousands of variants in a single experiment and have been shown to enable reclassification of 44% of VUS within each assayed gene. Leveraging these large and complex datasets, however, is challenging and time-consuming for variant scientists at genetic testing companies and diagnostic laboratories, who spend hours manually searching the literature and evaluating MAVE datasets in an ad hoc and error-prone process. There are no commercial platforms that curate the growing body of MAVEs and perform the rigorous analyses required to responsibly apply the data as evidence within classification guidelines established by the American College of Medical Genetics (ACMG) and the Association for Molecular Pathology (AMP). Constantiam’s technology reduces the time spent searching and evaluating MAVE data by 90%, removing a critical barrier by comprehensively curating MAVE data and transforming it into evidence that can be applied towards classification in minutes. During Phase I, Constantiam Biosciences successfully built a prototype of Varify, leveraging breakthroughs in Bayesian inference and probabilistic programming to analyze raw MAVE data for use as evidence. In parallel, we developed a user-facing application, MAVEvidence, for presenting data and generating evidence reports. The progress made in Phase I has laid a strong foundation for the Phase II project, which includes the following aims: 1) Expand Varify to include all MAVEs for cancer genes, and evaluate performance against clinically annotated variants; 2) Evaluate the functional evidence generated by Varify using population data from UK Biobank; and 3) Integrate Varify-derived evidence into MAVEvidence and develop automation capabilities for scalability. This proposed Phase II plan will enable Constantiam to complete development of a first-to-market, commercial-ready functional evidence platform. At the end of Phase II, genetic testing companies and diagnostic labs will be able to purchase Varify + MAVEvidence to reduce the time spent searching and evaluating functional data by 90%, addressing a $435M market opportunity while increasing the actionability of genetic tests for physicians and patients.
