Algorithms General Technology

Has DeepMind just cracked one of the biggest problems in Biology?

November 30, 2020

DeepMind isn’t just about teaching machine intelligence how to play games. Many of the projects the company is working on have very serious and very direct applications, especially when it comes down to biology.

It should come as no surprise then to many that DeepMind is often cooking up something in their AI kitchen that could directly impact the ability to solve protein structure – the so called “Protein Folding” problem which has confounded researchers and scientists for decades.

This computational work represents a stunning advance on the protein-folding problem, a 50-year-old grand challenge in biology. It has occurred decades before many people in the field would have predicted. It will be exciting to see the many ways in which it will fundamentally change biological research.


With their family of technologies prefixed with Alpha, we have seen AlphaGo and the ability to perform at the highest match level. But this time AlphaFold has out-performed around 100 researchers competing in the biennial protein-structure prediction challenge called CASP. CASP stands for Critical Assessment of Structure Prediction. One of the biggest problems biologists and researchers have faced in structural biology is understanding how genetic code (DNA, RNA) codes for the structures that comprise organisms (proteins).

AlphaFold had a median score of 92.5 out of 100, with 90 being the equivalent to experimental methods.

Comparisons of AlphaFold 2, previous attempts and other competitors.

Thrilled to announce our first major breakthrough in applying AI to a grand challenge in science. AlphaFold has been validated as a solution to the ‘protein folding problem’ & we hope it will have a big impact on disease understanding and drug discovery:

Demis Hassabis CEO of DEEPMIND on the recent development
Protein folding. Has one of the pinnacles of Biology been surmounted by the DeepMind team with AlphaFold 2?

AlphaFold’s astonishingly accurate models have allowed us to solve a protein structure we were stuck on for close to a decade, relaunching our effort to understand how signals are transmitted across cell membranes.

Professor Andrei Lupas, Director of the Max Planck Institute for Developmental Biology and a CASP assessor,

DeepMind will publish the work in due course, but the basic theme of the work and detail is covered by its past work on High Accuracy Protein Structure Prediction Using Deep Learning techniques. As yet none of those techniques employee Quantum Computing – but with developments in the field in the Quantum Machine learning and Quantum Neural Networks, there could be developments anytime soon which employ hybrid Quantum and Classical models.