Recently, we discussed the paper “Docking Screens for Novel Ligands Conferring New Biology” by Irwin and Shoichet. The paper discusses improvements in docking methods and demonstrates how docking campaigns can help to find new ligand series and new chemistries that can help reveal new biology.
The paper includes an extensive list of available docking resources: different web resources, compound libraries and docking software, together with the brief description of their application.
The authors highlight the danger that often comes with using docking as a black box – there is a great chance of the “garbage in, garbage out” effect where bad input results in bad output. Even a method that is (deceivingly) as simple as docking requires a team of experts that can thoroughly analyse the results and come up with several different potential lead series from the whole output rather than allowing the algorithm to pick the few best scoring ones by itself.
Another point was made that is often overlooked – there is an overwhelming amount of literature describing “docking for docking’s sake”. These are campaigns that result in good in silico hits, but these hits are then never confirmed experimentally. The authors briefly discuss the link between the docking and experimental artifacts and emphasise the need for orthogonal testing which, in our opinion, is rarely seen in publications dealing with any computational method.
We think this is a great summary of available docking resources that points out pitfalls associated with the approach that will inspire users to approach docking in a more thoughtful manner. However, we do feel that the title, although appealing, does not do it justice. The paper focuses on much more than just case studies where new chemotypes reveal new biology and it certainly is a necessary read for anyone embarking on a docking campaign for the first time or an experienced medicinal chemist.