SCOWLP: a web-based database for detailed characterization and visualization of protein interfaces

Abstract

Background: Currently there is a strong need for methods that help to obtain an accurate description of protein interfaces in order to be able to understand the principles that govern molecular recognition and protein function. Many of the recent efforts to computationally identify and characterize protein networks extract protein interaction information at atomic resolution from the PDB. However, they pay none or little attention to small protein ligands and solvent. They are key components and mediators of protein interactions and fundamental for a complete description of protein interfaces. Interactome profiling requires the development of computational tools to extract and analyze protein-protein, protein-ligand and detailed solvent interaction information from the PDB in an automatic and comparative fashion. Adding this information to the existing one on protein-protein interactions will allow us to better understand protein interaction networks and protein function.

Description: SCOWLP (Structural Characterization Of Water, Ligands and Proteins) is a user-friendly and publicly accessible web-based relational database for detailed characterization and visualization of the PDB protein interfaces. The SCOWLP database includes proteins, peptidic-ligands and interface water molecules as descriptors of protein interfaces. It contains currently 74,907 protein interfaces and 2,093,976 residue-residue interactions formed by 60,664 structural units (protein domains and peptidic-ligands) and their interacting solvent. The SCOWLP web-server allows detailed structural analysis and comparisons of protein interfaces at atomic level by text query of PDB codes and/or by navigating a SCOP-based tree. It includes a visualization tool to interactively display the interfaces and label interacting residues and interface solvent by atomic physicochemical properties. SCOWLP is automatically updated with every SCOP release.

Conclusion: SCOWLP enriches substantially the description of protein interfaces by adding detailed interface information of peptidic-ligands and solvent to the existing protein-protein interaction databases. SCOWLP may be of interest to many structural bioinformaticians. It provides a platform for automatic global mapping of protein interfaces at atomic level, representing a useful tool for classification of protein interfaces, protein binding comparative studies, reconstruction of protein complexes and understanding protein networks. The web-server with the database and its additional summary tables used for our analysis are available at http://www.scowlp.org.

Figure 1

Schematic overview of the methodology. SCOWLP uses information from PDB, SCOP, SCOL and SCOW for the computation of atomic interface interactions.

Figure 2

Comparative histogram of SCOWLP vs. PSIMAP database. Representation of the number of residue-residue (left y axis) and structural unit interactions (right y axis) contained in SCOWLP and comparison with PSIMAP.

Figure 3

Schematic representation of the interface interaction of two molecules and definition of wet spots. Molecules A and B form an interface. Interacting residues and water molecules are represented as black and open circles, respectively.

Figure 4

Enrichment of the interface definitions by peptidic-ligands and solvent. A) Enrichment in the description of protein interfaces by peptidic-ligands. The molecular recognition features of the BTB/POZ domain family are summarized. A representative POZ domain (green) is surrounded by five different ligands representing all possible BTB/POZ binding zones. Peptidic-ligands are represented in grey (PDB codes and chains used: 11dk_A, 11qb_ACD, 1r2b_D). B) Enrichment in the description of protein interfaces by wet spots. The complex of TEM1 β-lactamase (orange surface) with the inhibitor BLIPII (green ribbon; PDB 1jtd) is shown. White represents residues forming the interface before taking into account wet spots (in blue). C) The α and β (orange and green ribbon, respectively) chains of the Respiratory nitrate-reductase 1 (PDB 1q16) are shown. White represents residues forming the interface before taking into account wet spots (in blue). Figures created with InsightII, Accelrys.

Figure 5

Screenshots and legends showing the structure of the SCOWLP website.