Skip to Main Content
Data Science

Deep learning enables rapid identification of potent DDR1 kinase inhibitors

By Alex Zhavoronkov, Yan A. Ivanenkov, Alex Aliper et al

The process of drug discovery is cost-, labor-, and time-intensive, which constrains the ability to identify novel compounds with therapeutic efficacy to treat a variety of diseases. One potential way to enhance the compound-identification strategy is artificial intelligence (AI), specifically deep generative models; which use neural networks to produce objects with desired drug properties. Using a technique called generative tensorial reinforcement learning (GENTRL), potential compounds were generated for a molecular inhibitor of DDR1; a pro-inflammatory receptor tyrosine kinase involved in fibrosis and collagen deposition; inhibitors of which have therapeutic potential to treat diseases characterized by inflammation. Synthesis of 6 candidates generated using GENTRL were tested in vitro, demonstrated high selective activity against discoidin domain receptor 1 (DDR1) and discoidin domain receptor 2 (DDR2), and inhibited collagen expression and fibrosis. Using AI for drug discovery is a rapid (<2 months) and cost-effective approach to design effective drug candidates.

View the full peer-reviewed scientific paper at Nature.com.

Newsletter
Stay informed with our scientific views on healthy aging & longevity.
Why My Age Icon Why my age?
Yes, I would like to receive news, updates and product offers from Juvenescence by email. You can unsubscribe at any time. Privacy Notice
Success!
Welcome to your healthy aging community.