51 Tripeptides were synthesized and the zinc chelating activities were determined, which formed the database. Then those tripeptides were statistically analysed using 18 amino acid descriptors and partial least squares method. Results showed that the correlation coefficients of the quantitative structure-activity relationship(QSAR) models based on 3 amino acids descriptors met the requirements, which were the descrip-tor FASGAI, Z and C. And the descriptor FASGAI was the best with R2 = 0. 8225, Q2 = 0. 5818, RMSEE =0. 1628, Q2ext = 0. 6760, RMSEP = 0. 2499. Further analysis of the model on descriptor FASGAI revealed that the influence of amino acid position in polypeptide on the zinc chelating activity was N3>N2>N1 . Meanwhile, the bulky properties of amino acid residues influenced the zinc chelating activity of tripeptides.%人工化学合成了51条三肽,测定了其锌螯合活性,形成待测数据库.分别采用18种氨基酸描述符对三肽序列进行表征,利用偏最小二乘法进行统计分析.结果表明,3种氨基酸描述符对应的定量构效关系(QSAR)模型的相关系数达到建模要求,描述符分别为FASGAI,Z和C,其中描述符FASGAI最优,R2=0.8225,Q2=0.5818,RMSEE=0.1628,Q2 ext=0.6760,RMSEP=0.2499.进一步分析描述符FASGAI所构建的模型发现,在三肽序列中,氨基酸位置对三肽锌螯合活性的影响力大小顺序为N3>N2>N1,同时,多肽各位置上氨基酸残基的立体属性主要影响其螯合活性.
展开▼
