DeepMind’s AI will speed up drug discovery by predicting how proteins are weighed
Google DeepMind has developed a tool to predict the structure of proteins from their genetic sequence, marking a remarkable example of using AI in the process with scientific discovery.  How It Works: The system, called AlphaFold, models the complex folding patterns of long chains of amino acids, based on their chemical interactions, to form a three-dimensional form of a protein. This is known as the “protein week problem,” which has challenged researchers for decades.
Why It’s Important: The form of a protein dictates its function in the body so that predicting a protein’s structure allows researchers to synthesize new protein-based drugs to treat diseases or new enzymes to break down contaminants in our environment .
Training data: The DeepMind team trained deep neural networks to predict the distances between amino acids and angles between their chemical bonds, using the massive amounts of data available from genomic sequencing. The resulting system generates very accurate protein constructions that exceed previous prediction techniques, the team says.
The larger image: DeepMind is not the only one that works to speed up scientific detection with machine learning. Many other companies and researchers have tried to develop algorithms to detect new drugs and new materials.