Friday, August 12, 2022

AI model discovers drugs 1,000 times faster than traditional methods

- Advertisement -
- Advertisement -


Researchers have developed an artificial intelligence model that can find potential drug molecules more than 1,000 times faster than current state-of-the-art methods.

The Massachusetts Institute of Technology (MIT) team says the AI ​​model, dubbed EquiBind, will significantly reduce the chances and costs of drug trial failures.

The number of molecules with potential drug-like properties is gigantic and is estimated to be around 1060 For comparison: the Milky Way has about 108th Stars.

The EquiBind model is able to successfully bind these drug-like molecules to proteins at a rate 1,200 times faster than any of the fastest existing computational molecular docking models.

EquiBind achieves this through integrated geometric thinking, which allows predicting which proteins will match a molecule without prior knowledge of the target pocket.

“We were amazed that EquiBind was able to put it in the right pocket when other methods were completely wrong or only one right, so we were very happy to see the results for this,” said Hannes Stark, a first grader PhD student in the MIT Department of Electrical Engineering and Computer Science and lead author of the article describing the research.

The findings have already attracted the attention of industry insiders in hopes that they can be used to find treatments for lung cancer, leukemia and gastrointestinal tumors.

“EquiBind provides a unique solution to the docking problem that includes both pose prediction and binding site identification,” said Pat Walters, chief data officer for drug discovery firm Relay Therapeutics.

“This approach, using information from thousands of publicly available crystal structures, has the potential to impact the field in new ways.”

The paper, entitled EquiBind: Geometric Deep Learning for Drug Binding Structure Prediction, will be presented at the International Conference on Machine Learning (ICML).

- Advertisement -
Latest news
- Advertisement -
Related news
- Advertisement -

LEAVE A REPLY

Please enter your comment!
Please enter your name here