ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

Logos for amino-acid preferences in different backbone packing density regions of protein structural classes

Kannan, N and Schneider, Thomas D and Vishveshwara, S (2000) Logos for amino-acid preferences in different backbone packing density regions of protein structural classes. In: Acta Crystallographica Section D, 56 (9). pp. 1156-1165.

[img] PDF
Logos_for_amino-acid_preferences_in_different_backbone_packing_density_regions_of_protein_structural_classes.pdf
Restricted to Registered users only

Download (1265Kb) | Request a copy

Abstract

A protein sequence can be classified into one of four structural classes, namely $\alpha, \beta, \alpha + \beta \hspace{2mm} and \hspace{2mm} \alpha/\beta$, based on its amino-acid composition. The present study aims at understanding why a particular sequence with a given amino-acid composition should fold into a specific structural class. In order to answer this question, each amino acid in the protein sequence was classified to a particular neighbor density based on the number of spatial residues surrounding it within a distance of 6.5 \AA. Each of the four structural classes showed a unique preference of amino acids in each of the neighbor densities. Residues which show a high compositional bias in a structural class are also found to occur in high neighbor densities. This high compositional bias towards specific residues in the four different structural classes of proteins appears to be caused by structural and functional requirements. The distribution of amino acids in different neighbor densities is graphically presented in a novel logo form which incorporates several features such as composition, the frequency of occurrence and color code for amino acids. The spatial neighbors of the residues in different neighbor densities and their secondary structural location are also represented in the form of logos. This representation helped in the identification of specific details of the whole data which may otherwise have gone unnoticed. It is suggested that the data presented in this study may be useful in knowledge-based structure modelling and de novo protein design.

Item Type: Journal Article
Additional Information: Copyright of this article belongs to International Union of Crystallography.
Department/Centre: Division of Biological Sciences > Molecular Biophysics Unit
Date Deposited: 08 Oct 2007
Last Modified: 19 Sep 2010 04:40
URI: http://eprints.iisc.ernet.in/id/eprint/12149

Actions (login required)

View Item View Item