Abstract
Abstract The physiological functions of proteins are decided by the molecule structures. Therefore, analysis of protein structures plays an important role in modern biology research. The structure alignment is a powerful tool to compare and find similar proteins. Based on different criteria of different approaches, some protein structure alignment algorithms have been developed, and widely applied to protein structure analysis. The structure alignment method, SARST, which performs sequence alignment on the converted information of Ramachandran plots of proteins, has been proposed to provide fast alignment and searching abilities within a large protein structure database. In this thesis, we implement SARST and add some enhancements to improve it to be a more applicable software tool and available on Internet. SARST is tested with some real protein structure data from PDB, and analysis of parameters of SARST is included. Although SARST may not offer superior pairwise alignment qualities, it provides a much faster one-against-all searching for similar protein structures. Therefore, SARST can be a good filter to get fast rough searching, combined with other well-defined structure alignment tools as post-processing to get accurate final alignments.