Project 9. Charting the Space of Binding-Site Architectures in Proteins: A Large-Scale Analysis


Key project members: Tina Ritschel
Sponsor: Deutsche Forschungsgemeinschaft (DFG)
Drug discovery programs typically start with a high-throughput screen (HTS) for compounds exhibiting the desired biological activity. To maximize the number of hits, compound libraries screened by HTS are enriched with so-called privileged scaffolds, chemical substructures known to be present in ligands for multiple receptors. While the existence of privileged scaffolds is generally accepted, little is known about why these substructures are so special.
This study aims to systematically analyze the ligand binding sites of protein receptors to reveal the relationship between privileged scaffolds and their binding sites. To identify the characteristics of receptors that make them preferably interact with privileged scaffolds, a large scale analysis will be performed of all protein-ligand complexes deposited in the Protein Data Bank (PDB). To render this analysis computationally feasible a special binding-site representation, based on pharmacophore fingerprints, will be used.
The resulting chart of naturally occurring binding sites in protein space, will allow for the extraction of privileged binding site characteristics, which in turn will lead to improved design principles for screening compound library design.

Project Goals
The hypothesis, that there are some binding-site characteristics that make target proteins preferably interact with compounds containing privileged scaffolds, can be tested by using the increasing number of available protein-ligand complexes experimentally determined in atomic detail at the RCSB Protein Data Bank25) (PDB). In contrast to previous work we are not looking into a single protein family, rather the complete protein space is analyzed.
By using efficient methods to represent and compare binding sites this study will extend the scope of the analysis to process all currently known 3D-structures of proteins. Extraction of the binding sites from the PDB should reveal biases in binding site architectures. The extensive nature of the structure comparisons will allow us to draw general conclusions about binding sites characteristics in proteins that will eventually lead to the definition of privileged binding sites.

Perform a large-scale analysis of protein-binding-site architectures charting the space of naturally occurring binding pockets.
Identify privileged binding site characteristics that can lead to improved design principles for screening compound library design.
Wood, D.; de Vlieg, J.; Wagener, M.; Ritschel, T. Pharmacophore fingerprint-based approach to binding site subpocket similarity and its application to bioisostere replacement. J. Chem. Inf. Model, 2012, 52, 2031-2043.