DeepMind publishes structural predictions for SARS-CoV-2

DeepMind publishes structural predictions for SARS-CoV-2

New York, 9.3.2020

To help the global research community better understand the coronavirus, DeepMind today released the structural predictions for six proteins associated with SARS-CoV-2, the virus that causes COVID-19, and the most recent version of them used AlphaFold system (released).

As the world struggles with the COVID 19 outbreak, one research team after another has grown in the global scientific community to offer expertise, tools and possible solutions. In the early stages of the outbreak, front-line laboratories were open-source genomes of the virus that allowed other researchers to quickly develop tests around the pathogen. Other laboratories modeled the coronavirus infection peak or produced molecular structures to develop drug compounds and treatments for infections.

Understanding the structure of a protein is an important resource for understanding how a virus works. However, experiments to determine the structure usually take months or longer. To speed things up, researchers have developed calculation methods to predict the protein structure from amino acid sequences.

In January, DeepMind released AlphaFold, a deep learning system designed to accurately predict protein structure even when no structures of similar proteins are available, and to generate 3D models of proteins with SOTA accuracy. According to DeepMind, AlphaFold has been improved since release to get more accurate predictions. The results of the predicted structures for some of the proteins in SARS-CoV-2 that were generated using their newly developed methods are now available to the public.

„We have confirmed that our system provides accurate prediction of the experimentally determined SARS-CoV-2 spike protein structure that is shared in the protein database,“ wrote DeepMind researchers on their official blog. „We recently shared our results with several colleagues at the Francis Crick Institute in the UK, including structural biologists and virologists who have encouraged us to now make our structures accessible to the general scientific community.“

Although current structural predictions have not yet been peer-reviewed or experimentally verified due to the seriousness and timing of the situation, DeepMind has decided to release the predicted structures now, in the hope that the work can help survey the scientific community, such as virus functions and provide a platform to hypothesize future experimental work in the development of treatments.

The structural predictions, relevant technical details and data can be downloaded here.