Application of a deductive database system PACADE toward discovery of clusters of similar structures in proteins

We developed the deductive database system PACADE (Protein Coordinate Analyzer with Deductive Engine) for analyzing three dimensional and secondary structures of protein. PACADE has an inference system and a relational database which stores primary, secondary and tertiary structures converted from Brookhaven's Protein Data Bank (PDB). Users of PACADE can represent biological knowledge on structures of protein, in a logical and declarative form and some iterative structures can be efficiently searched for, using PACADE. However, for protein structure analysis, any function that can only search for structures described by users is inadequate and a function was designated to search for similar structure in proteins. Using this function, similar structures can be searched for without calculating the inter-atomic root mean square distance. The outline of the PACADE system and the function and some results of searches for super-secondary structure, using the function are described, in detail. With this mode, one may be able to discover clusters of proteins based on similarity of structure.<
>