Abstract
Central dogma of molecular biology states that DNA carries the genetic information which is transcribed to mRNA and subsequently translated to protein. And function of protein is determined from protein structure. Widespread researches all use the sequence alignment to join the structure information to find out the latent function. This method often causes not alike or relationship of evolutionary the farther sequence can't find out the function. According to this, the usage of structure alignment builds up evolutionary tree in this thesis, by the structure alignment tree and the sequence alignment tree to the relation of that analyze the sequence, structure, function and evolution. The first step of our method used carries on the alignment to the protein structure through the CE tool and will need of as a result save the distance matrix. Using the UPGMA algorithm come to carries on to the distance matrix to divide into the cluster and the usage of evolutionary tree in proper order to present. For each protein structure, we join the EC number that it belongs to mark clearly its function. The dataset that this thesis uses is the TIM-Barrel fold under the SCOP classification. Discard the mutants and will have after many PDB files of other multi-domains carry on incising. According to the SCOP most the first floor classification Species picks 239 protein structures random respectively to carry on structure alignment. After noting through the establishment of the structure alignment tree and the mark of the EC number, will it with ClustalW produce of the sequence alignment tree does the analysis comparison. By as a result know, in spite of the sequence identity of high or low. The function cluster for structure alignment tree is more consistent and can find in same function cluster rather far sequence of existing the relation of sequence evolutionary. even can see accordingly belong to the different superfamily but exist wait for various sequence, structure function and evolution in same function cluster. According to this reason, this research suggests while carrying on taking structure as the sequence alignment of the base. Also can use the structure alignment to as a result do the mutual analysis and find out some latent relations by this.