BioPredict, Inc.             
Home Company Profile Technologies Services Internal Discovery Success Stories People Contact



Fragment Based Drug Discovery

Fragment-based methods have become a key core technology in the arsenal of structure-based drug discovery methods.   In particular, small-molecular-weight molecules called fragments have proven extremely useful in identifying chemical motifs to target multiple targets within protein families and superfamilies.  A key point in this approach is that the probability that a chosen ligand fits an active site is inversely proportional to its size, since larger ligands are more likely to be more specific and therefore less likely to fit any given active site.  This imparts an advantage to initially conducting a smaller, less specific fragment screen, especially when targeting multiple members of a structural family.  While initial fragment hits bind with low affinity, follow-on steps use these fragments  (i)  to search for larger compounds that share critical pharmacophoric features, (ii) as scaffolds for combinatorial chemistry library synthesis, or (iii)  to generate larger higher affinity leads by linking several fragments together.  Advantages of the fragment based method include (i) elucidation of hits using extremely small screens, (ii) the existence of a structural hypothesis for every hit, and  (iii)  the ability to use these structural hypotheses to rapidly identify larger molecules with higher affinities or to generate family-focused combinatorial libraries.   Theoretically there is a valuable entropic gain to assembling larger ligands from fragment that individually fit an active site. This gain can manifest as several orders of magnitude in binding constant.

  The success of fragment-based methods relies on selection of initial compounds using structural hypotheses, followed by elaboration of experimentally verified hits into larger molecules to achieve desired levels of activity and target selectivity.  When initial structural conjectures prove correct the “hit to lead” and “lead optimization” interval of drug discovery process can be shortened with a higher probability of success. 

When approaching a target protein family or protein we will often apply fragment-based approaches either alone or in parallel with other methods to search for hits.  We have assembled an internal virtual library of ~40,000 purchasable small molecules/fragments that satisfy  criteria which represent a distillation of our own and others experience, and incorporate the “Rule-of-Three” discussed by Congreve [Congreve].

Virtual screens of fragments are conducted and analyzed using  a combination of commercial software and our own internally developed hypothesis-driven docking software, pharmacophore mapping software, and post-docking analysis software  (see Technologies).

The use of fragment-based discovery is illustrated in our approach to the GHKL superfamily of proteins, described elsewhere on this website (see Internal  Discovery and success stories).  

[Congreve] Congreve M, Carr R, Murray C, Jhoti H.  A 'rule of three' for fragment-based lead discovery? Drug Discovery Today. 2003, 8:876-7.