DeepMind has an new artificial intelligence that can predict genetic diseases
How to Run with the Missense Mutation? The Puzzle of Predicting Genomic Disorders and Diseases: The Case for iga Avsec
iga Avsec was a PhD physics student when he found himself taking a machine learning course at the university. He was soon working in a lab that studied rare diseases and the aim was to pin down the exact genetic cause of the disease.
Many of the genetic disorders, such as cystic fibrosis and sickle cell disease, are caused by genetic alterations that change the sequence of the molecule that the disease is caused by. A tiny number of these single-letter missense mutation have been observed. Of the more than 70 million possible in the human genome, only a sliver have been conclusively linked to disease, and most seem to have no ill effect on health.
In spite of the millions of missense variations found in humans, only 2 percent have been categorized as either benign or harmful by years of research. A single missense variant can take a long time to study.
Pushmeet Kohli, DeepMind’s vice president of research, uses the analogy of a recipe book. If AlphaFold was concerned about the binding of ingredients,AlphaMissense predicts what would happen if you use the wrong ingredient.
AlphaFold ushered in a new era of computational biology, according to Joseph Marsh, a computational biologist at theMRC Human Genetics Unit. It is exciting. It’s probably the best predictor we have right now. But will it be the best predictor in two or three years? There’s a good chance it won’t be.”
AlphaMissense can be applied in real-world settings if it is rigorously evaluated, using good performance metrics.
For example, an exercise called the Critical Assessment of Genome Interpretation (CAGI) has benchmarked the performance of such prediction methods for years against experimental data that has not yet been released. “It’s my worst nightmare to think of a doctor taking a prediction and running with it, as if it’s a real thing, without evaluation by entities such as CAGI,” Bromberg adds.
