Fragment-Based Drug Discovery

The principle in using fragments relies on the premise that these molecular entities are more efficient ligands compared to drug-like molecules, and their structures can be further optimized more efficiently. As small entities, molecular fragments can be iteratively optimized to show a better pharmacokinetic profile in the later development stages. Drug-like molecules may contain functional groups that contribute poorly to protein binding or, in some cases, can even disrupt the protein-ligand interaction. On the other hand, fragments often form high-quality interactions at the protein target and can bind to several locations translating to a greater number of hits.

Fragment growing is the most common strategy in FBDD due to its very nature as it requires only one fragment to bind somewhere at a target structure. On a given fragment decorations are added to increase the size and introduce addional interaction features. It is very similar to a classic lead optimization process where a ring moiety, a sidechain, or a functional group is added to elucidate structure-activity relationships.
Fragment linking on the other hand introduces a connection between two binding fragments. The simultanious binding of both fragments at the same time has synergistic effects on the potency of the linked compound, usually increasing the binding affinity manyfold compared to the binding the individual fragments.
The merging of fragments requires two fragments that share an overlap in a target's subpocket. The overlapping groups are fused together to connect the rest of the fragments into a single molecule.

All of the mentioned methods can be applied in SeeSAR. In the Inspirator Mode it is possible to perform fragment growing, linking, and merging. The library for growing can be easily modified by users to create a custom version with their own building blocks. Furthermore, users can explore the binding pocket by themselves within the Molecule Editor Mode and test their hypothesis with the visual feedback they receive from SeeSAR.
Several examples of FBDD with SeeSAR are available as tutorials.