Alan Cheng, Amgen Cambridge Research Center
I will present computational approaches for assessing and addressing selectivity and druggability issues in drug discovery projects. Selectivity analysis is important because unwanted secondary pharmacology is often an issue. We have developed approaches based on discriminating physiochemical features of binding sites. Druggability assessment is important because estimates are that 90% of the human genome is undruggable, and it would be valuable to identify the less-druggable targets beforehand to help prioritize research resources. Here, we have developed a biophysical approach that estimates the maximal affinity of a binding site for a drug-like molecule. We validated our approach through both retrospective data analysis and a prospective high-throughput screening experiment.