DeepMind’s gigantic leap: Solving 50 years old quest of biology

Complex of bacteria-infecting viral proteins modeled in CASP 13. The complex contains four separate subunits that were modeled individually. PROTEIN DATA BANK

DeepMind, an AI technology company offshoot of Google AI, has already amazed the scientific community showcasing AIs that have learned to play a variety of complex games with superhuman skill, from Go and StarCraft to Atari’s entire back catalog. It has now made an astronomical leap in solving previously unsolvable biological challenge. Determining the protein’s structure through its amino acid sequence.

Christian Anfinsen, 1972 Nobel laureate in Chemistry has postulated the theory that a Protein’s amino acid sequence should fully determine its structure. It sparked a long 50 years swedge to correctly model protein structure computationally based on one-dimensional amino-sequence. Which took a half-century to correctly 3D model protein structure because theoretically protein folding takes astronomical random shifts before synthesizing into a final 3D structure.

In 1969, Levinthal's paradox, a thought experiment which estimated 10^300 possible conformations for a typical protein if modeled using brute force technique probably take the age of the known universe to correctly configure all protein’s structure.

We have been stuck on this one problem — how do proteins fold up — for nearly 50 years. To see DeepMind produce a…