Last year, a group of researchers from Google Research Applied Science and X-Chem, published an article that describes an effective approach for hit finding using Machine Learning (ML) based on DNA-Encoded Libraries (DEL). The authors demonstrate how ML could be applied to DEL selection data by identifying active molecules from large libraries of commercial and tangible compounds. The researchers train models using only DEL selection data and apply special filters to the predictions across three diverse protein targets: sEH (a hydrolase), ERα (a nuclear receptor), and c-KIT (a kinase). As a result, an overall hit rate is around 30% and disclosed potent compounds (with nM activity) are dissimilar to the original DEL.
Chemspace inspire You to start new projects using DEL and we offer:
1. Blocks for creating DELs:
In-Stock Bifunctional Cores for DEL - 38 733 compounds
In-Stock Trifunctional Cores for DEL - 1 308 compounds
In-Stock Decorators for DEL - 119 467 compounds.
2. Ready to use self-serve product from HitGen Inc. - OpenDEL®:
“OpenDEL® - Billions in A Vial!” - 1 Billion compound set.
In case of any questions just drop us a line: firstname.lastname@example.org.