I just returned from the 2014 Workshop on Free Energy Calculations in Drug Design at Vertex Pharmaceuticals in Boston.
This meeting was interesting for me due to the number of attendees from industry---I would guess about half. I have not interacted with pharma much, so it was interesting to hear both "stories from the trenches" and thoughts about the role of free energy calculations in drug discovery.
While there was certainly plenty of skepticism about the role of free energy calculations in drug design, there was also guarded optimism. With the advent of GPU computing, these calculations are fast enough (or they will be soon) to provide answers on the timelines needed in industry. The usual question remain about force field accuracy and how to achieve sufficient sampling, but those will undoubtedly improve in the future.
Michael Shirts suggested that some of our most pressing problems are human in nature. How do we know that our codes are implemented correctly? Do we have enough testing? Are our software development practices solid enough? How do we ensure that practitioners are following best practices? What are best practices?
Overall, it was a great meeting. The next edition will be in 2016.