Abstract
Finding Consistent Sequence patterns in Glycoside Hydrolases (GH) Protein Families Chin-Liang Tsai, Advisor: Professor Chuan Yi Tang Master of Science on Computer Science, National Tsing Hua University, Hsinchu City, Taiwan In this research, our major objective is finding the consistent sequence patterns in GH protein families. GH sequences can be classified into 125 families by the sequence similarities. The PROSITE databases only build 18 motif descriptions or patterns for GH families. Some multiple sequence alignment (MSA) tools cannot handle a large of sequences in addition. We classified GH proteins by sequence identity and randomly selected one sequence from every cluster to build their patterns. We compared with using all sequence to build and our method. The sensitivity of our patterns is greater than using all sequence. The patterns were consistent when we randomly selected sequences. We used this method to construct the patterns for GH family proteins. Key words: Glycoside hydrolases, Multiple sequence alignment, Patten