S-Space College of Natural Sciences (자연과학대학) Program in Bioinformatics (협동과정-생물정보학전공) Theses (Ph.D. / Sc.D._협동과정-생물정보학전공)
Development of Globularity and Torsion Based Analysis Methods for Protein Structures
구형성과 뒤틀림각에 기반한 단백질 구조 방법론 개발
- 자연과학대학 협동과정 생물정보학전공
- Issue Date
- 서울대학교 대학원
- protein structure; protein folding; structural globularity; torsion angle system; Levinthal paradox; cotranslational folding
- 학위논문 (박사)-- 서울대학교 대학원 : 협동과정 생물정보학전공, 2013. 2. 손현석.
- The structure of protein has intimate relationship with the function of protein. The structure of protein is experimentally determined through X-ray crystallography and NMR methods. However, X-ray crystallography is hard to obtain mobile protein structure and crystallization often causes practical problems. NMR structure is impossible in the observation of membranous or large proteins. Thus, theoretical methods for the determination of protein structures are highly concerned to circumvent practical problems. Homology, threading and ab initio modeling are the three typical approaches in protein structure modeling. ab initio modeling is often called as protein folding problem. The natural stable state of protein structure is believed to be the minimal energy state. The critical problem of protein folding research is the impossibility of the exhaustive search of possible conformations. Globularity of the protein structure was assessed in the pursuit of the universal structural constraint while approximated measurement name Gb-index was developed. Strong perfect globe-like character and the relationship between small size and the loss of globular structure was found among 7131 proteins which implies that living organisms have mechanisms to aid folding into the globular structure to reduce irreversible aggregation. This also implies the possible mechanisms of diseases caused by protein aggregation, including some forms of trinucleotide repeat expansion-mediated diseases. Torsion angle constraint mimics natural process of conformational change of proteins which lacks significant movement along covalent bonds and change in bond angles. This torsion angle system was applied to structure alignment to prove the validity as a structural representation. It was more effective to accurately anticipate homology among 1891 pairs of proteins of 62 different proteases and among 1770 pairs of 60 proteins of kinases and proteases with the string of φ and ψ dihedral angle array than famous 3D structural alignment tool TM-align. Secondary structure database and structure alignment web server was constructed from PDB and SCOP entries based on the simple classification scheme according to the backbone torsion angles. The database introduced here offers functions of secondary database searching, secondary structure calculation, and pair-wise protein structure comparison. Visualization during the process of the protein folding simulation is quite interesting regarding the fast apprehension of the states while previous algorithms such as molecular dynamics offers very few options of interference. Computational application named ProtTorter which visualizes three-dimensional conformation, calculates the potential energy, and supplies the user interface for backbone torsion angle manipulation was developed. Using this application, simple folding algorithm was newly investigated. Cotranslational and torsional folding path was utilized in the context of Levinthal paradox. The validity of the folding method was investigated using the test sets of small peptides. Positive result for the possibility of this method was obtained as the stable negative energy minimal structures and fast convergence. Application of torsional system of which validity was proved in the structure alignment assays and globular constraints which might infer solvent interactions by minimizing solvent accessible surface area might be worth for further studies based on the folding algorithm using ProtTorter application.