DeepMind has predicted the structure of almost every protein known to science

0
15

However, for many proteins, “we are interested in how their structure is changed by mutations and natural allelic variation, and that will not be covered by this database,” AlQuraishi said. “But of course the field is developing rapidly and so I expect tools to accurately model protein variants will emerge soon,” he added.

The quality of AlphaFold’s predictions may also not be as accurate for rarer proteins with less available evolutionary information, Peng says.

The move is the latest development in DeepMind’s push into “digital biology,” where “AI and computational methods can help understand and model key biological processes,” Hassabis said. Hassabis also leads a new venture, also owned by Alphabet, called Isomorphic Labs, which is developing AI for drug discovery.

Pushmeet Kohli, chief of AI for science at DeepMind, said the company has many challenges in the life sciences it still wants to address, such as how proteins behave and interact with other proteins.

Hassabis said his dream is that AI can not only help figure out the structure of proteins, but also become “an important part of the discovery process for new drugs and cures.”