SIOC English
化学学报
高级检索

化学学报 ›› 2006, Vol. 64 ›› Issue (5): 393-396. 上一篇    下一篇

研究论文

一组新氨基酸描述子用于肽定量构效关系研究

梁桂兆1,2,3，周鹏1，周原1,2,3，张巧霞1，李志良*,1,2   

  1. (1重庆大学化学化工学院 重庆 400044)
    (2湖南大学化学生物传感与计量学国家重点实验室 长沙 410082)
    (3重庆大学生物工程学院 重庆 400044)
  • 投稿日期:2005-06-15 修回日期:2005-11-03 发布日期:2006-03-15
  • 通讯作者: 李志良

New Descriptors of Aminoacids and Their Applications to Peptide Quantitative Structure-Activity Relationship

LIANG Gui-Zhao1,2,3, ZHOU Peng1, ZHOU Yuan1,2,3, ZHANG Qiao-Xia1, LI Zhi-Liang*,1,2   

  1. (1 College of Chemistry and Chemical Engineering, Chongqing University, Chongqing 400044)
    (2 State Key Laboratory of Chemo/Biosensing and Chemometrics at Hunan University, Changsha 410082)
    (3 College of Bioengineering, Chongqing University, Chongqing 400044)
  • Received:2005-06-15 Revised:2005-11-03 Published:2006-03-15
  • Contact: LI Zhi-Liang

PDF

939

被引次数

61 | 11

用主成分分析从20种天然氨基酸0D～3D结构信息中收集到的共1369个描述子变量得到了一组新氨基酸描述子(SZOTT), 将其用于血管紧张素转化酶抑制剂和苦味二肽结构表征并以偏最小二乘法建立定量构效关系模型, 得复相关系数RCU2分别为0.894和0.908, 留一法交互检验的复相关系数RCV2分别为0.828和0.736, 估计均方根误差RMS分别为0.331和0.195. 研究结果表明, SZOTT描述子含信息量大, 操作简便, 结构表达能力强, 有望在多肽定量构效关系研究中得到进一步推广.

关键词: 肽, 定量构效关系, 主成分分析, 遗传算法, 偏最小二乘法

A new set of descriptors, namely vector of scores for zero dimension, one dimension, two dimension and three dimension (SZOTT), was derived from principle component analysis of a matrix of 1369 structural variables of 0D～3D information for 20 coded aminoacids. SZOTT scales were then employed to express structures of angiotensin-converting enzyme inhibitors and bitter tasting thresholds, and to construct QSAR models based on partial least square (PLS). The results obtained are as follows: the multiple correlation coefficient (RCU2) of 0.894 and 0.908, the leave one out cross validated RCV2 of 0.828 and 0.736, and root-mean-square error for estimated error (RMS) of 0.331 and 0.195, respectively. Satisfactory results showed that, as new aminoacid scales, data of SZOTT may be a useful structural expression methodology for study on peptide QSAR (quantitative structure-activity relationship) due to its many advantages such as plentiful structural information, easy manipulation, and high characterization competence.

Key words: peptide, quantitative structure-activity relationship, principal component analysis, genetic algorithm, partial least square