Date Published: March 01, 2017
Publisher: International Union of Crystallography
Author(s): Charlotte M. Deane, Ian D. Wall, Darren V. S. Green, Brian D. Marsden, Anthony R. Bradley.
The background to and development of WONKA and OOMMPPAA, tools for structure-based drug design, are described.
Technological advances in high-throughput crystallography and protein–ligand biophysical and biochemical binding assays have resulted in a rapid increase in the quantity of liganded crystal structures and high-quality activity data points for many protein targets (Badger, 2012 ▸; Zheng et al., 2014 ▸). Concerted efforts to consolidate and store such data have generated large and highly curated data sets both in the private (for example corporate databases) and public domains (Berman et al., 2003 ▸; Gaulton et al., 2012 ▸). Further, it is now commonplace for an industry structure-based drug-design (SBDD) programme to have access to many tens of liganded crystal structures and many thousands of high-quality activity data points.
A full description of both the WONKA and OOMMPPAA methods can be found in separate publications [Bradley et al. (2015) ▸ and Bradley et al. (2014) ▸, respectively]. Data are stored in a bespoke Python Django (Django Software Foundation, 2013 ▸) data model that is common to both applications. All computational chemistry processing was carried out using RDKit (Landrum; http://www.rdkit.org). The input data for both tools are pre-aligned PDB files of protein–ligand complexes. A comma-separated variable (CSV) file is required to indicate the path to the PDB file, and the SMILES (Weininger, 1988 ▸) specification is required for the ligand bound to that protein. Additionally, for OOMMPPAA activity data are required and are input as a separate CSV file. Here, we give a brief overview of the methods.
In this paper, we describe the background for and use of the WONKA and OOMMPPAA platforms. Both methods are freely available interactive computational tools designed to analyse and describe the influx of protein–ligand interaction data associated with SBDD programmes. WONKA is a tool to summarize large ensembles of protein–ligand structures of the same protein target. WONKA also provides a platform for annotation and data sharing within and between research groups, a feature that is invaluable in the context of working in a multi-disciplinary team. OOMMPPAA builds upon WONKA to incorporate available activity data in the context of the binding sites of protein–ligand structures using a 3D MMP approach. Further, we show the use of OOMMPPAA in interrogating available activity data for smaller (BRD4) and larger (carbonic anhydrase 2) structural and activity data sets. Both WONKA and OOMMPPAAA are freely available to try online and are free to download at http://oommppaa.sgc.ox.ac.uk/OOMMPPAA/ and http://wonka.sgc.ox.ac.uk/WONKA/.