This is a post that introduces the idea of molecular complexity and FBLD. In 2001, Hann and colleagues GSK developed the theory of molecular complexity, an idea that largely revolves around the principle that simpler molecules can be accommodated in a variety of protein sites compared to their more complex counterparts. The fact that molecular simplicity is inherently linked to molecular size meant that the tenet of molecular complexity could provide a likely explanation for successes within the fragment-based lead discovery (FBLD) paradigm. Fragments are defined in a few different ways, of which “half the size of a drug” is quite a popular one. Since Lipinski’s famous rule of five sets the bar for the molecular weight of a drug molecule at 500 Da, fragments can be considered as molecules with molecular masses ≤ 250 Da. Use of the FBLD approach is often based on three primary advantages:
(i) Following on from the molecular complexity idea, smaller and inherently simpler fragments, should be less prone to steric mismatches within protein binding sites, providing higher success rates when it comes to hit identification. This becomes starker when dealing with challenging sites such as protein-protein interfaces, where hit-rates from screens of drug-sized molecules may even approach zero.
(ii) At the Lipinski limit of 500 Da, there are thought to be more molecules (drug-like) possible than there are stars in the universe, or at the very least, an astronomical number of molecules. This ensures that even the largest of compound collections, which often get populated with molecules of this size, covers but a miniscule fraction of chemical space. At fragment-size (250 Da), the vastness of chemical space shrinks rapidly, allowing even small libraries to explore a much greater fraction of this space.
(iii) The molecular complexity idea also states that a complex molecule that “fits” ideally into a binding site is likely to possess high affinity, target selectivity, and a single binding mode. Once fragment scaffolds are identified, these can be “grown” to an optimum level of molecular complexity taking into account ADMET properties, CNS-related properties, multi-target and selectivity requirements. This makes the approach ideally suited to handle complex tasks in early-stage drug discovery involving challenging targets with difficult disease profiles.
The name “molecular jigsaws” is based on applying the elegant molecular complexity idea to achieve lead compounds with the desired characteristics for each project. In other words, we consider each project a complicated puzzle for which the appropriate jigsaws are created and brought together using an array of computational techniques. We of course always strive for a beautiful picture when the pieces come together!