Abstract

The concept that compounds with reduced molecular complexity sample chemical space more efficiently than larger molecules is central to Fragment-Based Drug Discovery. At Astex, our fragment library has progressively evolved to lower average molecular weight, driven by an analysis of the impact of compound size on historical hit rates, and enabled by the high sensitivity of X-ray crystallography as a primary screening technique. A logical extension of this might be to examine the effects of further reductions in ligand size, to a level which approaches discrete functional group probes. This talk will describe how we have explored this through the construction and crystallographic screening of a dedicated chemically diverse library of ultra-low molecular weight fragments or MiniFrags.

I will present results showing that high concentration soaks in this size regime give a significant increase in binding events compared to conventional fragment screening. I will also show that MiniFrags provide a more complete energetic mapping of already precedented binding sites by highlighting additional points for favourable interaction. These “warm-spots” support weaker binding than the “hot-spots” detected using standard screening protocols, but can nevertheless be productively engaged during medicinal chemistry campaigns to guide the design and optimization of fragment-derived compounds.