Modeling the decisions of a medicinal chemist
What is it about?
We used a medicinal chemists scores of over 300 chemical probes to build machine learning models. We validated the models and compared to other metrics for scoring molecules.
Why is it important?
To our knowledge this was the first time anyone has take this strategy and prospectively tested such models. We show that we can take a medicinal chemists decisions as an input alongside chemical structure and help prioritize other compounds. This could be useful for selecting chemical probes that are medicinal chemistry friendly.
The following have contributed to this page: Dr Sean Ekins, Dr Nadia Litterman, and Dr Christopher A Lipinski