Alpha Missense: Proteome-Wide Missense Variant Effect Prediction

Explore a groundbreaking conference talk from the Broad Institute's Models, Inference and Algorithms series, featuring Jun Cheng from Google DeepMind discussing Alpha Missense. Delve into the innovative adaptation of AlphaFold for predicting missense variant pathogenicity in the human genome. Learn how this model combines structural context and evolutionary conservation to achieve state-of-the-art results across genetic and experimental benchmarks. Discover the model's ability to predict cell essentiality and identify short essential genes that elude traditional statistical approaches. Gain insights into the comprehensive database of predictions for all possible human single amino acid substitutions, classifying 89% of missense variants as likely benign or pathogenic. The talk is preceded by a primer from Žiga Avsec, providing essential background information on the topic.

MIA: Jun Cheng, Alpha Missense; Primer by Žiga Avsec