Conformational Epitope Prediction based on Geometrical Relationship and Antigenic Characteristics
| Autor: | Wei-Chun Chung, 鐘煒鈞 |
|---|---|
| Rok vydání: | 2008 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 96 Conformational Epitopes (CE) are composed of discontinuous stretches but closely distributed in three-dimensional spaces which play an important role in the antigen-antibody binding. The successful prediction of conformational epitopes is of great importance in disease diagnosis and vaccine design. The main goal of this thesis is to develop an effective and efficient system for prediction of CEs of a resolved protein structure. The proposed system is the first tool that adopts three-dimensional mathematical morphology to define protein structural surface regions,. Based on the surface information, the spatial distance relationship and the antigenic characteristics are considered as key features for CEs candidate identification. In this thesis we have adopted 33 antigen-antibody complexes with verified conformational epitopes for the evaluation of the prediction accuracy, and the results were compared with the present prediction tool PEPOP. To observe the influence of chemical property scales and surface ratios, we have performed a complete evaluation by various parameter settings. It has been discovered that the sensitivity was superiorly improved by selecting hydrophilicity feature better than hydrophobicity, while the specificity favor using hydrophobicity feature in stead of hydrophilicity in general. In this thesis, the results have shown an overall accuracy of 75% can be approached, and this system possesses comparable performance with respect to the existing prediction tools. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|