HELP

PB-PENTAPEPT is a platform which combines a database and a couple of tools to investigate pentapeptides in protein structures. PB-PENTAdb provides a database which provides different backbone structural diversity features of pentapeptides in protein structural domains. Potential applications of this database are presented here as tools for the prediction of local backbone structures (PB-kPRED) and their inherent structural variability (PB-SVindex).

Pentapeptides are analyzed here in terms of a structural alphabet which are unique, short local structures that are recurrent in protein structures but also in terms of secondary structure (from DSSP) and solvent accessibility (NACCESS). We used the Protein Blocks (PBs) which is the most widely used structural alphabet in computational biology to date (Karchin,R., et al, Proteins, 51:504-514, 2003 ; Joseph A.P. et al, Biophys. Rev., 2:137-145, 2010).

PBs are a set of 16 pentapeptide structural motifs (de Brevern et al, Proteins 41:271-287, 2000). They are abstraction of the unique local backbone structures present in proteins and upon which the backbone model of the proteins can be built. Each of the 16 protein blocks is defined by a vector of 8 dihedral angles associated with five consecutive residues and represented by an alphabet character from a to p (see picture below).

This ability to abstract protein 3D structures into 1D sequences of PBs (PBE-T) lead to the development of new approaches for protein structure analysis (Offmann et al, Current Bioinformatics, 2:165-202, 2007 ; Joseph A.P. et al, Biophys. Rev., 2:137-145, 2010).

In the same line, web services were proposed by us earlier: PBE (Tyagi et al, Nucl. Acids. Res., 34:W119-123, 2006) available at PBE and more recently, iPBa (“iPBA: a tool for protein structure comparison using sequence alignment strategies”, Gelly J-C et al, Nucl. Acids. Res., in press, 2011) available at http://www.dsimb.inserm.fr/dsimb_tools/ipba/.