Is the Force with Us? Quantifying Model Uncertainty in Molecular Simulations

Most drug discovery projects have a 90% failure rate. In this talk I will show how I used PyMC3, a Python probabilistic programing language, to improve molecular models in the early stages of drug discovery. I will also discuss how I made the data I generated easily reusable to extend the chemical space of the models in the future.

 Speaker: Chaya Stern, Weill Cornell Medicine / Memorial Sloan Kettering Cancer Center
 Speaker: Gregory Ross, Schrodinger
 Speaker: John Chodera, Memorial Sloan Kettering Cancer Center