FAMILY SPECIFIC ANALYSIS OF BETA BULGES & BETA TWIST IN BETA SANDWICH PROTEINS

蛋白质的折叠结构以不同的方式描述

The folded structure of a protein is described at different levels: primary structure, secondary structure, tertiary structure and quaternary structure. Beta-Sheets are made of several strands and must be listed as tertiary structural elements. Naturally found beta-sheets are not flat, but rather display stronger or weaker left-handed rotation (viewed along an axis lying across the strands in the sheet). Further, irregularities in the hydrogen-bonding pattern, so-called "beta-bulges", are not uncommon. There are today enough protein families with several determined 3D-structures to enable an analysis of geometrical parameters describing the twist and beta-bulges within families of homologous proteins (i.e. evolutionarily related through a common ancestor). A database of known "beta-sandwich" structures has been derived from the available 3D- structures (Brookhaven Data Base). We are currently analyzing and compiling various geometrical parameters (e.g. inter-strand and intra-strand angles) to describe the beta-sheet twist properties and occurring beta-bulges in the sandwich database structures. We will use the data first to examine the relation between the distortion from a regular sheet (i.e. with constant, average twist parameters) of pairs of homologous proteins and their evolutionary distance. By correlation of the geometrical data with the corresponding mu ltiple sequence alignments, the results of this study should allow me to develop new heuristics to improve protein structure prediction from amino acid sequence data. For example, many of the commonly encountered difficulties with predicting beta-strand secondary structure arise from insufficient knowledge about the beta-twist and beta-bulges in the sheet structures. The Computer Graphics Laboratory resources provide the means for computing and visualizing the geometrical parameters that are surveyed in the project outlined above, and further the means to analyze the correlation of these parameters with primary sequence data.