Google AI’s DeepMind AlphaFold now accurately predicts the structure of nearly every protein known to science

AlphaFold 2 has halved the error rate for protein structure prediction to a level where it can correctly predict almost every protein structure in the human body. The ability to predict the structure of a protein from its genetic sequence is one of the grand challenges in biology, as knowing the structure of a protein gives vital clues to its function and can lead to the development of new drugs. The AlphaFold system developed by Google-owned AI company DeepMind has been shown to be highly accurate in predicting the structure of proteins, and in 2021 it was awarded the prestigious Nature Methods Method of the Year Award. In the latest version of AlphaFold, published in Nature, researchers have further improved the accuracy of the system, making it possible to predict the structure of nearly every protein known to science. The research team tested AlphaFold on a set of 208 proteins that had been experimentally determined to high accuracy. AlphaFold was able to correctly predict the structure of 187 of these proteins, with an average error of just 0.6 angstroms. This is a significant improvement over the previous version of AlphaFold, which had an average error of 1.2 angstroms. The researchers also showed that AlphaFold can accurately predict the structure of proteins that are much larger than those previously studied. They were able to correctly predict the structure of a protein complex that contained 33 subunits, which is the largest protein complex that has been successfully predicted by AlphaFold. The improved accuracy of AlphaFold is likely to have a major impact on biology and medicine. The ability to accurately predict the structure of a protein can help researchers to understand how proteins work, and can lead to the development of new drugs and therapies. For example, AlphaFold has already been used to identify new targets for drugs that could treat Alzheimer’s disease and Parkinson’s disease. The researchers behind AlphaFold say that they hope the system will eventually be used to predict the structure of every protein in the human body. This would be a major milestone in biology, and could lead to a new era of drug discovery and personalized medicine..