1910s promising AI model inked in the JCIM

Announcement

The first-ever AI-familiar biotech groundbreaking large and small molecule therapeutics discovery, 1910, has proudly announced that its robust AI model, CANDID-CNS™, is now a part of the Journal of Chemical Information and Modeling (JCIM). Being published in an American Chemical Society journal proves 1910’s continuous efforts and hardship to introduce AI to the healthcare sector.

The publication’s headline is “CANDID-CNS™: AI Unlocks Stereochemistry and Beyond Rule of 5 to Predict CNS Penetration of Small Molecules”. This AI model is the first-ever model that on-point predicts the blood-brain barrier (BBB) penetration for Beyond Rule of 5 (bRo5) molecules and comprises stereochemistry, as these are the two main hurdles in neuroscience drug discovery.

Published data

The CANDID-CNS™ identifies the toughest issues that come across the drug discovery by analysing which molecules can meet the blood-brain barrier. The BBB stops almost 100% of large molecules and above 98% of small molecules from entering the CNS space. This makes neuroscience the toughest therapeutic space in pharma R&D. Most of the approved CNS drugs are small molecules that align with Lipinski’s Rule of 5.

Whereas, the bRo5 compounds are crucial and huge molecules that mark an untapped chemical class with noteworthy potential, uncovering the undruggable targets. The bRo5 molecules face three major challenges, such as the inability of previous computational methods, difficulty in penetrating the BBB and exclusion from several medicinal chemistry designs. The stereochemistry swayed BBB permeability, but the latest computational models didn’t mention this. CANDID-CNS™ has successfully dealt with these limitations by precisely revising stereochemical distinctions and predicting BBB permeability for bRo5 molecules that administer CNS penetration.

The PhD, Founder and CEO of 1910, Jen Asher, said, “The Neuroscience definition has been out of reach for us. CANDID-CNS™ extends the boundaries of the considered druggable in the brain. By overcoming these limitations of learning stereochemical impact and bRo5 design, it has opened the doors for a new bRo5 chemical area for CNS drug discovery. This has brought us one step closer to excellent, effective treatments for ALS, Alzheimer’s and Parkinson’s diseases.”

The PhD, lead author of the JCIM publication and Senior AI Research Scientist at 1910, Jesse Collins, said, “CANDID-CNS™ not only justify the molecules in a classified form but also recovers the physicochemical principles that determine BBB transport. Its predictions connect with the quantum mechanical hydration free energy, confirming that the model implicitly thoroughly learns the thermodynamic determinants of passive permeability.”