Events

Computationally Enabled Drug Discovery to Switch the Immune System Off and On: From Physics to AI/ML

Since our inception in 2009, Nimbus Therapeutics has been a leader in computationally enabled drug discovery. In our first decade, we successfully applied physics-based software such as FEP+ and Water Map to predict binding affinities of small molecules to enzymes. By combining these physics-based programs with high throughput structural-biology, we optimized potency against desired targets while designing exquisite selectivity over antitargets. Additionally, we used physics-based methods to predict properties such as cell permeability and solubility to more rapidly optimize our molecules. Over the past five years, however, we have increasingly been incorporating artificial intelligence/machine learning (AI/ML) for both potency and property predictions to build best-in-class drugs. Through selective incorporation of AI/ML, we have augmented our computational toolbox in multiple ways to improve property predictions and to accelerate high quality potency predictions. This presentation will provide a Nimbus historical perspective on the application of physics-based tools to discover a drug for the treatment of autoimmune and immunology diseases (i.e. turning the immune system off) and the discovery of immuno-oncology agent NDI-101150, an HPK1 inhibitor for the treatment of cancer (i.e. turning the immune system on). Lastly, I will describe our successful incorporation of AI/ML strategies to accelerate the identification of potent HPGD inhibitors for the treatment of inflammatory bowel disease (IBD).

Read more about Nimbus Therapeutics here