Analysis of three dimensional structure of membrane proteins using bioinformatics and computational analysis.

摘要

Transmembrane (TM) proteins are estimated to represent ∼20-30% of the genome and are of great cellular and therapeutical significance. Functioning as receptors, transporters or enzymes, membrane proteins mediate a broad range of fundamental cellular activities. Structure determination of such proteins by experimental methods is very challenging. As a result only a small number of structures are reported. Bioinformatics and computational approaches have proven to be very useful in structure prediction of these proteins. In this thesis we have used these approaches in three different aspects to analyze the 3-dimensional structure of membrane proteins.;For structure prediction of membrane proteins, developing an accurate scoring function for structure discrimination and validation of membrane proteins remains a challenge. Network approaches based on the overall network pattern of residue packing have been proven useful in developing scoring function for soluble protein structure discrimination. This approach has been adopted for membrane proteins. The derived scoring function acts as good indicator of a native fold, and is very effective for discriminating less native membrane protein folds from native ones.;Elucidating the distinct topology of residue packing in transmembrane proteins is essential for developing high-quality methods for their structure prediction. Along with the network approach, evolutionary conservation of the protein residues and their solvent accessibility was applied on a non-redundant dataset developed for membrane proteins and soluble proteins, in order to understand the packing arrangement of the important residues. Together with the knowledge of a centralized function site for many membrane proteins, a distinct model for membrane proteins was suggested.;Understanding the structural basis for the ligand-binding selectivity of the seven helical membrane proteins is of significance to their structure prediction. Comparison analysis of proteins' ligand binding sites provides a useful way to study their structure-activity relationships. A pairwise comparison study of the binding sites was done using the seven helical membrane protein structures including the bacteriorhodopsin and the G-protein coupled receptors. This study provided novel insight into the structural basis of ligand-binding selectivity of seven-helical membrane proteins, and is of practical use to the computational modeling of these proteins.