Quantitative prediction of class I MHC/epitope binding affinity using QSAR modeling derived from amino acid structural information.
Quantitative prediction of class I MHC/epitope binding affinity using QSAR modeling derived from amino acid structural information.
复制标题
使用源自氨基酸结构信息的 QSAR 模型定量预测 I 类 MHC/表位结合亲和力。
DOI：
10.2174/1386207318666150121125746
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
Lin, Zhihua
中科院分区：
文献类型：
--
作者：
Shu, Mao;Hu, Yong;Xia, Qingyou;Lin, Zhihua
T细胞受体(TCR)识别T细胞免疫应答，T细胞免疫应答依赖于TAP转运的多肽抗原与主要组织相容性复合体(MHC)分子结合。MHC-表位结合亲和力的定量预测可以方便表位筛选，大大减少成本和实验工作量。本研究利用氨基酸的物理化学信息建立了定量构效关系(QSAR)模型，建立了结合亲和力的综合定量预测方法。首先，通过一系列氨基酸物理化学参数对表位进行了表征。其次，采用逐步回归方法对结构变量进行优化。最后，对31个MHC I类亚型进行多元线性回归(MLR)，建立了稳健的定量模型。QSAR模型的归一化回归系数(NRCS)可以很好地解释MHC、表位和TCR的相互作用机制。NRC计算的表位每个位置的氨基酸贡献可以决定哪一个更有利于结合亲和力。因此，STR-MLR所建立的定量模型可用于指导CTL表位的虚拟组合设计和高通量筛选。此外，它们还具有物化指标明确、计算和解释容易、性能好等优点。
The activation of T cell immune responses, which relies on peptide antigens transported by TAP and bound to major histocompatibility complex (MHC) molecules, is recognized by T cell receptors (TCR). The quantitative prediction of MHC-epitope binding affinity can facilitate epitope screening and reduce cost and experimental efforts greatly. In this study, a comprehensive quantitative prediction method of binding affinity was established using quantitative structureactivity relationship (QSAR) modeling derived from amino acid physicochemical information. Firstly, the epitope was characterized by a set of amino acid physicochemical parameters. Secondly, the structural variables were optimized by the stepwise regression (STR). Finally, the robust quantitative models with were built by multiple linear regressions (MLR) for 31 MHC Class I subtypes. The normalized regression coefficients (NRCs) of QSAR model could demonstrate the mechanism of interaction of MHC, epitope, and TCR very well. The contribution of amino acid at each position of epitope, which was calculated by NRC, could determine which one was favorable for binding affinity or not. Therefore, the quantitative models established by STR-MLR could be used to guide virtual combinational design and high throughout screening of CTL epitope. Besides, they have many advantages, such as definite physiochemical indication, easier calculation and explanation, and good performances.