The University of Tokyo and Google Japan have signed an agreement to collaborate on AI-based Polygenic Risk Score (PRS) model research. This collaboration is part of the partnership between the University of Tokyo and Google, which was concluded in October 2020, with the aim of solving various social problems and creating a future society in which AI and humans can coexist in harmony.

Google has previously developed AI models to improve the accuracy of variant calling, the process of identifying mutations and variants 1F 0F 2 in genome sequence data, and published in Natur e Biotechnology in 2018. We are also developing models that can more accurately predict physical characteristics such as eye color or specific diseases from genome sequence data and use this to discover variants associated with specific characteristics. The Institute of Medical Science at the University of Tokyo has made the largest contribution in non-Western populations to numerous genome-wide studies of obesity, blood tests, type 2 diabetes, myocardial infarction, atrial fibrillation, and other diseases through analysis of data accumulated in Biobank Japan.

Press Release (Japanese only)