I wrote this awhile back….. Hard to believe it was almost 2.5 years ago when I had just left medical school and started my biomedical engineering masters when I wrote this. I would agree with most of what was said here today except that I would argue that knowledge based algorithms based on forcefields and physical chemistry would be more trusted than machine learning approaches when trying to estimate energy.

AI and Drug Discovery

The world of drug discovery is seeing a paradigm shift due to artificial intelligence. Chemoinformatic approaches that combine chemistry with programming are now forming an integral part of a drug discovery pipeline (Faulon et al, 2010). Algorithms can be used to make predictions for target discovery, toxicity, and binding efficacy, among others, before doing lab testing. While not a substitute for lab experiments, chemoinformatic approaches are useful as they allow screening an enormous chemical space, reducing costs and aiding in hypothesis generation (Faulon et al, 155).

Virtual Screens

Searching for new targets is an expensive endeavor. In order to bring a new drug to market, it can take 10–15 years and cost 1 billion dollars. Drug development costs so much because of the science. Biology is complex, and 95% of drug candidates fail. Scientific American outlines three reasons drug discovery is so difficult: