【摘要】 从20种天然氨基酸的50个物化性质出发,按照疏水、立体和电性特征将其分类后分别进行主成分分析,并将产生的得分矢量即VHSE(principalcomponentscorevectorofhydrophilicity,steric,andelectronicproperties)作为氨基酸结构描述子用于肽的定量构效关系研究。与已有方法相比,VHSE描述子具有物化意义明确、结果更易解释等特点。应用该描述子并结合逐步回归变量筛选和偏最小二乘建模方法,在对苦味二肽和血管舒缓激肽促进剂等体系的定量构效关系研究中,均取得了优于已有文献的结果。更多还原

【Abstract】 Peptides are of central importance in all living systems. They have attracted considerable pharmacological interest in recent years. For peptide properties a precise amino acid sequence is required for a particular function i.e. activity. A quantitative structure-activity relationship (QSAR) will then indicate how the change in peptide sequence, is correlated with the change in biological activity. In this work, a new set of descriptors, VHSE (Principal component score vector of hydrophilicity, steric and electronic properties), was derived from principal components analyses of 50 physicochemical variables of 20 natural amino acids separately according to different kinds of properties described, namely, hydrophilicity, steric, and electronic properties. As a new set of AA scales, VHSE is of definite physical and chemical meaning and easy interpretation. The scales were then applied in QSARs of two sets of peptides using partial least squares regression on top of stepwise regression. Better squared multiple correlation coefficient, R2, and cross-validated R2, namely Q2, of the resulting QSAR models were obtained in comparison with those obtained with other 2-D or 3-D descriptors.更多还原