Binding Site Prediction

Binding Site Prediction for Rational Drug Discovery

At the beginning of every drug discovery or lead optimization campaign is the question about the potential binding site of a ligand.
Independent if a binding site is orthosteric or allosteric, collecting ideas about interactions between a target structure and the ligand is essential. These insights form the foundation to decide what compounds to synthesize next, helping to validate the hypothesis about the binding mode and to guide the design novel and improved candidates.

This page elaborates on what to keep in mind during the binding site prediction and how to form a hypothesis for the next steps.

How to Predict a Binding Site at a Protein

During the folding of the protein, solvent exposed cavities form. Those potential binding sites can accommodate smaller molecules or parts of larger structures and are ultimately used as spots to modify the activity of a target. Upon binding, ligands stabilize particular conformations of the target altering its behaviour.
Binding sites can be formed entirely by the target (e.g., protein, DNA, RNA, ...), or once it engages with other partners forming a tertiary complex.

Therefore, understanding the structure of the target protein is essential, as its conformation heavily influences the shape and characteristics of its binding sites. Examining multiple PDB entries can help identify a consensus on key interaction points, providing valuable insights for targeting valid binding sites effectively.

Finding Hot Spots

SeeSAR's Binding Site Mode helps you to detect potential binding sites of your target structure based on their predicted druggability.
To better assess the properties of the binding sites, relevant data, such as the size of the cavity and its lipophilicity, are provided for your analysis. The calculations are swift and can be applied on single proteins, functional complexes, oligomers, as well as DNA and RNA.

BioSolveIT software for binding site prediction:

SeeSAR: Visual, drug design dashboard for computational and medicinal chemists.
Identify potential binding sites with a single click in SeeSAR's Binding Site Mode.

Command-line tools for binding site prediction:

DoGSiteScorer: Automated pocket detection.

Validating Your Hypothesis

To validate your prediction, it’s essential to combine information on ligand activity with insights into its interactions with the target structure.
Dock your known actives into the predicted binding site and assess the predicted binding modes to rationalize structure-activity relationships (SARs). Having a series of compounds to work with significantly simplifies the process.

Important functionalities of the compounds should bind in areas where they are tolerated or, ideally, form favorable interactions. Our scoring algorithm HYDE helps you spot the individual contributions of each heavy atom, as well as the quality of the formed H-bonds. Build up on gained insights to iterate your initial hypothesis and adjust accordingly.

Additional Considerations

Take into consideration which residues of the potential binding site are conserved. Highly conserved residues of a target or protein class usually play a crucial role in the receptor functionality. Conserved residues in the orthosteric binding site of a protein family sharing the same endogenous ligand may engage with it through interactions.
Create close analogs to your compound by adding, removing or replacing parts of the molecule or its heavy atoms to confirm your binding site prediction.
Compound Series (in-house or publicly available) provide you will a lot of data points to model your hypothesis with.
Consider to take water molecules, as well as co-complexed ligands into account if they play a role for the binding of the ligand.
Last but not least, site-directed mutagenesis is a powerful tool to narrow down your binding mode. Combine it with SAR insights to valuable insights.

Excited for more drug discovery solutions?

Keep on exploring!

Binding Site Prediction