| 加权相关系数核函数及其在T细胞表位预测中的应用 |
| |
| 作者姓名： | Huang J Ma JH Liu N Qian SS |
| |
| 作者单位： | 南方医科大学生物医学工程学院,广东,广州,510515 |
| |
| 基金项目： | 南方医科大学博士启动基金,国家自然科学基金 |
| |
| 摘 要： | T细胞表位的识别和理论预测是计算机辅助疫苗和药物设计中的重要研究内容,在药物设计、疫苗设计、以及帮助人们了解机体免疫功能方面具有重要的指导意义.本文源于多肽的"锚点特征",设计一种加权相关系数,并作为支持向量机的相似函数核,结合演化算法,以支持向量机为基础对T细胞表位进行了理论预测.在同样编码方式下,实验表明采用本文设计的加权互相关系数核函数,其支持向量机模型的预测性能显著提高.
|
| 关 键 词： | T细胞表位 核方法 支持向量机 遗传算法 交叉相关系数 |
Improved weighted cross-correlation coefficient with a new kernel and its application in predicting T cell epitopes |
| |
| Huang J,Ma JH,Liu N,Qian SS.Improved weighted cross-correlation coefficient with a new kernel and its application in predicting T cell epitopes[J].Journal of Southern Medical University,2010,30(10):2252-2255. |
| |
| Authors: | Huang Jing Ma Jian-hua Liu Nan Qian Shan-shan |
| |
| Institution: | School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China. hjing@fimmu.com |
| |
| Abstract: | We designed a weighted cross-correlation coefficient considering the "anchor" of the T cell epitopes, and used an evolutionary algorithm to search for an optimal weight vector. A SVM model with this new peptide similarity kernel was evaluated on a T-cell data set. The results demonstrated a good performance of this method. |
| |
| Keywords: | |
| 本文献已被 万方数据 PubMed 等数据库收录！ |
|
