Health

Researchers Match DeepMind’s AlphaFold2 Protein Folding Power with Faster, Freely Available Model

Researchers Match DeepMind’s AlphaFold2 Protein Folding Power with Faster, Freely Available Model

Deepmind shocked the biological world late last year when its Alphafold 2 AI model predicted the structure of proteins (a common and very difficult problem) so accurately that many declared the decades-old problem a “solution.” Now researchers claim that the way Deepmind has leapfrogged Deepmind to other parts of the world, RoseTFFold, is a system that does almost the same thing at a fraction of the calculation cost. (Oh, and it’s free to use). Alphafold 2 has been a topic of discussion in the industry since November, when it eliminated competition in CASP14, a virtual competition between algorithms designed to predict the sequence of amino acids in the physical structure of proteins.

Deepmind’s model was far ahead of the others, so high and reliably accurate, that many in the field talked about moving into a new field (with half-importance and good humor). One aspect that no one seemed to be satisfied with, however, was that Deep Mind’s system plan for it was not fully and publicly described, and some people expressed concern that the company (which is owned by Alphabet / Google) planned to keep the secret sauce – Which will be their pioneer but will also be against the principle of mutual assistance in the scientific world.

This concern seems to have spread at least somewhat from the research work of researchers at the University of Washington, led by David Baker and Minkyung Beck, published in a recent issue of the Science Journal. Baker, you may recall, recently won a breakthrough award for his team’s work in the fight against COVID-19 with engineered proteins. The team’s new model, the RoseTTFold, predicts similar levels of accuracy using the method of answering questions through the imam, inspired by the alpha fold 2 recognized by Mether in the same way.

“At the CASP14 meeting, the Alphafold 2 group introduced a number of new high-level concepts. Minking has been able to make amazing progress in a very short period of time, starting with these ideas and with a lot of teammates working together in the brain, “he said. (” He’s amazing!”He added.)