Fragment growing

Why it’s important to get the right lead molecule

Research has shown that candidate molecules tend to be closely related to the lead compounds, therefore, the potential success of a new drug campaign is highly dependent on the selection of lead series very early on in the drug discovery process.

Challenges in hit to lead stage

Drug discovery is a costly and time-consuming process with the cost of the hit to lead stage alone estimated at 166 million USD. Identifying drug-like molecules that target significant metabolic pathways plays a crucial role in the ability of producing new medicines, however, because of the dramatic increase in the available molecule information, screening approaches have obvious limitations in search space coverage.

Hit discovery technologies range from traditional high-throughput screening to  fragment-based techniques, affinity selection of large libraries and computer-aided de novo design. Development of computational fragment-based approaches has contributed to the acceleration of the drug discovery process by facilitating the screening of libraries of compounds and reducing the pool of compounds for synthesis. Nostrum Biodiscovery has been contributing to those efforts by developing our own fragment growing tool.

Fragment growing

FragPELE can automatically grow hundreds of fragment molecules onto a docked ligand scaffold in a high-throughput manner. At each step, it runs multiple PELE simulations to efficiently sample the re-arrangement of the system as the fragment is grown. The protocol for growing ligands consists of a number of distinct steps:

  1. Fragment linkage. At this stage, a fragment is covalently linked to the ligand core at a position defined by the user.
  2. Fragment reduction. The parameters of the fragment atoms are reduced to later be regrown inside the binding site.
  3. Fragment growing. During a series of steps, the fragment is grown, iteratively increasing its parameters.
  4. Sampling and scoring. In the last step, a PELE simulation is performed to score the grown molecule based on the interaction energy between the ligand and the protein.

The combination of Monte Carlo sampling with the growing algorithm used in FragPELE allows the complex to adapt while exploring the significant areas of the potential energy surface.

FragPELE interactins

Interaction diagrams of the ligand with the protein before the fragment growing (left) and after the fragment growing (right). The growth of the fragment and the adaptation of the complex optimize the binding interactions.

Research at Nostrum Biodiscovery

As a company specializing in drug design, we understand the importance of seemingly insignificant structural changes and the vast impact they can have on binding affinity, which is why we have been improving FragPELE to tackle some of the most common challenges in drug design.

Water displacement

The new implementation can be used to predict the principal hydration sites or the rearrangement and displacement of conserved water molecules upon the binding of a ligand, letting the users explore the resulting changes in free energy.

Library screening

Additionally, we’ve been working on an automated workflow to make your everyday hit to lead studies more bearable! You will soon be able to screen thousands of compounds from public libraries, filter them according to your requirements and assess their binding affinity.

Feed Forward Frag

Finally, we are developing a new workflow to iteratively perform fragment growing simulations and extract the top candidates from an external dataset based on similarity to identify the most promising lead compounds.

FragPELE: Dynamic Ligand Growing within a Binding Site