Abstract
It is a useful strategy to understand structure-function relationship of proteins by aligning and analyzing their sequences. Database search techniques that can extract the functional or structural information from amino acid sequence are urgently needed. Imidase superfamily provides an interesting case for developing such method. Imidase super family contains a variety of related proteins that differ in functions and sequence identities. We developed a Cross-Reference Voting algorithm with complexity O(nm 2 ), where n is the number of total sequences and m is the maximum length of the sequence, to identify amino acid candidates that are critical for function of rat imidase. After analyzing more than 80 members of imidase superfamily, several amino acid residues of rat imidase are selected and examined. His67, His69, Lys159, His192, His248 and Asp326 of rat imidase are found corresponding to critical residues of hydantoinase studied by site-directed mutagenesis and structure information of hydantoinase. His67, His69, Asp100, His217, His248, His420 and His459 of rat imidase are found corresponding to critical residues of dihydroorotase and allantoinase in separate studies. These results indicated that our algorithm could successfully extract functionally related amino acid candidates in imidase superfamily.