Deargen optimizes prediction of candidate drug structures

By The Science Advisory Board staff writers

April 9, 2021 -- Artificial intelligence (AI) drug discovery and development company Deargen is touting results of its controlled molecular generator (CMG) technology that overcomes limitations associated with existing models of predicting molecule properties.

The firm presented its findings during the ACM Conference on Health, Inference, and Learning (ACM CHIL), held April 8-9. Deargen said its new CMG technology improved performance by nearly double compared with existing models such as molecule deep Q-networks (MolDQN) and variational junction tree encoder-decoder (VJTNN).

The firm tested aniracetam, which has the weakest binding affinity for dopamine D2-type receptor (DRD2), among 28 DRD2-targeted compounds. The researchers found DRD2 binding affinity was enhanced with nearly no change in other properties of aniracetam through the use of Deargen's CMG. Deargen's technology can optimize various molecular properties simultaneously and can aid in the development of new drugs, as it takes less time to analyze those properties, the firm said.

Copyright © 2021

Science Advisory Board on LinkedIn
Science Advisory Board on Facebook
Science Advisory Board on Twitter