Online Molecular Docking Resources

Introduction

In the actual period we assist to the expansion of structural data concerning proteins and protein–ligand complexes with direct implications in the development of the computational strategies to enhance drug discovery. One of these strategies is the virtual screening (VS) technique that allows hits identification with lower costs compared with the traditional experimental high-throughput screening (Gohlke & Klebe, 2002). Virtual screening techniques can be classified into ligand-based and structure-based methods respectively.

The ligand-based method is used when a set of ligand molecules is known but the target’s structure is unknown and it may be implemented using several approaches. When the information contained in the set of known ligands is used to build a model of the receptor molecule, the ligand-based method is implemented as pharmacophore modeling (Sun, 2008). The ligand-based approach may be implemented using the quantitative structure activity relationship (QSAR) method that takes into account the connection between chemical structures of some chemical compounds and their biological activities and predicts the activities of new chemicals (Nantasenamat et al, 2009). Another approach is the chemical similarity searching, a method based on the assumption that all compounds in a database that are similar to a query compound have similar biological activity (Willet et al., 1998). When applying this method, a database of molecules is screened against one or more known active ligands structures in order to predict their biological activity. The ligand-based virtual screening may also be based on the shape similarity searching, molecules with shape similar to that of the active compound are considered to fit the target's binding site (Rush et al., 2005).

The structure-based drug design involves molecular docking as the most common method. It is based on docking of a candidate compound into the protein target and applying a scoring function to estimate the probability that the ligand will bind to the protein with high affinity (Kroemer, 2007). Computational programs based on diverse algorithms were developed to implement molecular docking studies, molecular docking being an increasingly important tool in actual pharmaceutical researches. From this point of view, it becomes important to evaluate the existing docking programs in order to choose the most suitable docking algorithm for a particular study.

Molecular docking studies may be implemented by using commercial software, free software designed for academic users and also using on-line specialized servers. There are many servers allowing online protein-protein and protein-ligand molecular docking. For example, Click2Drug server (http://www.click2drug.org/) contains a comprehensive list of docking web services, databases and computer-aided drug design tools.

Within this chapter, the use of some online resources designed to protein modeling and online servers allowing molecular docking studies based on different approaches is presented. The description is illustrated with examples of molecular docking computations concerning the prediction of the interactions of the cytochrome P450 with the fungicide difenoconazole (DFC). The results obtained using different servers are compared and discussed in terms of agreement or disagreement concerning the predictions of the receptor-ligand complexes obtained using the docking tools.