Protein structural information management based on spatial concepts and active trigger rules

Abstract

A protein structure has four different levels of structures with spatial conformation and arrangement of peptide chains. Features of newly emerging protein structure data are extremely complex, large, multidimensional and incomplete. Therefore, it is necessary to develop approaches to manage protein structure data compared with management of conventional data in order to provide protein structural information for analysis applications. We propose a new approach to manage protein structural information by using spatial object management and active database techniques. We introduce data modeling for protein structures using a geographic object model and version management of a protein sequence with trigger rules and show analysis structural information using topological and geometric operators. Our experimental results show that applying spatial models and an active trigger to manage protein structure can efficiently support fast protein structure analysis by using spatial operations and a filter-refinement processing with a multi dimensional index. © Springer-Verlag Berlin Heidelberg 2003.