- 學位論文
利用支援向量機與類神經網路預測抗原決定位
Immunogenicity Prediction with Neural Network and Support Vector Machine
摘要
抗原被抗體辨識或結合的地方叫抗原決定位（antigenic determinant或epitope）。抗原上的決定位通常含6～8個胺基酸，它可以是三度空間的構造(conformation structure)，通常是抗體結合到蛋白質暴露在外面的位置。T細胞所辨認的抗原決定位，它會與主要組織相容抗原（major histocompatibility complex，MHC）上的classI / II結合，並與T細胞上的T細胞受器（T cell receptor；TCR）結合，一個抗原通常有好幾個抗原決定位，構造越複雜，分子量越大，它的抗原決定位愈多。主要組織相容抗原的專一性可以由病原菌蛋白質胺基酸排列的順序進行預測，如選擇特定的抗原決定位並用於預防注射，協助了解發生感染性疾病、自體免疫疾病和癌症時免疫系統所扮演的角色。 本論文是根據抗原蛋白胺基酸的物理化學性質，利用類神經網路及支援向量機來預測抗原決定位。
關鍵字
抗原決定位並列摘要
Epitope, also named as antigenic determinant is the region that is recognized by antibody. In general, the length of the region in a protein sequence is about 6 to 8 amino acids. It is located on the exposed part in the protein structure for the reason that binding to the antibody. T cell epitopes are bound to MHC I/II on the surface of antigen-presenting cell and also bound to the receptor of T cell in the same time. One antigen could contain many epitopes, the number of epitopes increase with the increasing of the structural complexity and the molecular weight in the antigen. Epitope prediction is one of the important issues in vaccine design. In this study, we applied neural network and support vector machine to predict epitomes based on physiochemical properties of amino acids.