Extraction of patterns in each domain of G-protein-coupled receptors

Jeongho Huh1, Chungoo Park2, Dong Soo Jung, Hong Gil Nam, Jiin Choi, Young Bock Lee
1artist3@postech.ac.kr, Division of Molecular Life Sciences, Pohang University of Science and Technology; 2madreach@bric.postech.ac.kr, Biological Research Information Center(BRIC), Pohang University of Science and Technology

G-protein-coupled receptors (GPCRs) constitute the largest known family of cell-surface receptors, and they have enormous significance for the pharmaceutical industry. Their functional identification and classification is the first step in drug target discovery. Since many receptors remain with little structural information available, similarity search using programs such as BLAST and FASTA is one of the widely used methods in this step. However this method is limited in GPCRs analyses because many GPCRs have low sequence homology in a same functional group. Therefore, in this case, the detection method of local functional sequence patterns is more suitable. For this reason, we attempted to extract patterns that are specific in GPCR subtypes. Regardless of their low sequence homology, GPCRs have a common structure. This consists of three distinct topological domains that have different functions and secondary structures. From this biological information, we applied different rules to each domain for efficient pattern extraction. The concept of amino acid clustering was applied to extracellular and intracellular domains, and on the other hand, the structural feature of a-helix sequence was adopted to transmembrane domain. Consequently, we obtained specific patterns of GPCR clans with high frequency and high specificity. This collection of patterns is able to provide sensitive diagnostic opportunities for assigning GPCR subtypes.