Google is using AI to help scientists understand diseases (blog.google)

Google has announced significant advancements in its health research initiatives, deploying AI tools to enhance our understanding and treatment of various diseases. Over the past five years, Google teams have published 90 peer-reviewed papers and established collaborations with clinicians and researchers, exemplified by the creation of the AlphaGenome Atlas, which catalogs genetic variants and their effects on human biology. One practical application saw researchers using AlphaFold to identify two FDA-approved drugs for repurposing against Chagas disease, pointing to AI's role in accelerating drug discovery. The initiative also includes pioneering tools like DeepVariant and DeepConsensus for rapid genetic diagnosis in time-sensitive cases, achieving genome sequencing in under eight hours—an unprecedented feat acknowledged by Guinness World Records. Other groundbreaking projects involve ultrarapid detection of pediatric leukemia mutations and predictive modeling of global flu trends, which outperformed multiple forecasting systems. By harnessing machine learning, Google aims to streamline medical solutions and support healthcare providers, ultimately driving improved health outcomes and interventions worldwide.