DNA-encoded library screening combined with machine learning (DEL-ML) is a promising approach to hit-finding in early drug discovery, but the field is still working out which strategies for data representation, model architecture, and hit selection actually deliver results.
The first DREAM Target 2035 DEL-ML Challenge was designed to explore exactly that. Teams of AI and computational chemistry experts worked to identify hits for the target protein WDR91, both retrospectively and prospectively, deploying a wide range of computational approaches. The results offer some useful lessons: the choice of training data emerged as the clearest differentiating factor between teams, while model architecture and data representation had less impact than might be expected. Novel WDR91 ligands were also discovered in the process.
This webinar will present an overview of the challenge and its results, followed by presentations from the top-performing teams. Registration is free but mandatory.
Chair: Matthieu Shapira, Structural Genomics Consortium
Speakers:
Luca Chiesa, Structural Genomics Consortium
Lei Zhang, New York University Shanghai
Yifan Jiang, The Hospital for Sick Children (Toronto) & Vector Institute
Anna Khapeliukha, Chemspace LLC
Jeff Messer, GSK
Wim Dehaen, University of Chemistry and Technology Prague
Register here to receive the Zoom link prior to the event.
