STAT Online November 30, 2020 “DeepMind’s protein-folding AI stuns with a solution to one of biology’s biggest challenges” By Casey Ross
See STAT article and Nature article for details.
Image from Nature.com
“At the start of a biennial contest to predict the structure of proteins, the expectations for Google’s artificial intelligence unit DeepMind couldn’t have been higher.” DeepMind demonstrated that it had conquered one of biology’s great challenges that is being able to “quickly and accurately” predict the three dimensional (3D) “structure of a protein from its amino acid sequence.” This breakthrough will speed innovation in any field studying proteins and then applying native or altered proteins to improve processes including health. With such understanding researchers can “better target those proteins to counteract mechanisms underlying disease.”
DeepMind’s program called “AlphaFold” outperformed mostly academic laboratories “by a wide margin, scoring about 90 out of 100. Experts are impressed but realize there’s still work to be done as not every submitted sequence is correctly solved. The performance is evidence that Google has learned to use AI for this application to solve many structures much quicker than existing methods that require “months or years.” Andrei Lupas (Max Planck Institute) noted that “DeepMind was able to predict the structure of a protein that had confounded his lab for more than a decade.” He went on to remark “It will change research. It will change bioengineering, It will change everything.”
Google plans to publish in a peer-reviewed journal but the AlphaFold team blogged that this is one of Google’s “most significant advances to date.” They noted “AlphaFold demonstrate[s] the stunning potential for AI as a tool to aid fundamental discovery.”