Development of fragments can be difficult and time consuming when there are no clear protein features to make new interaction and improve weak affinity fragments into higher affinity leads. Astex(1) has recently discussed a new approach to identifying hot spots and new potential interactions within proteins by screening small compounds of 5 to 8 heavy atoms called “minifrags”. Minifrags are screened through high concentration x-ray crystallography techniques (soaking ~ 1 M compound) in the presence of a lead fragment and can identify new expansion vectors and minimal pharmacophores ready to be linked or merged with the existing fragment to rapidly improve affinity.
There are numerous approaches to the design and development of compound libraries for screening. Here we analysed fragmented FDA approved, orally available drugs found in the DrugBank(2), commercially available space and the most commonly crystallised minifrags found in the Protein Data Bank(3) (1,4). These databases were analysed to provide guidance on the most common minifrags and total minifrag space to aid design of a minifrag library. The current presentation will describe the analysis of minifrag space, chemoinformatic design of the screening library, as well as the preliminary practicality of crystallisation at extremely high concentrations.