【摘要】 多肽保留时间预测有助于提高多肽鉴别的可信度以及增加正确鉴别的数目。本文利用氨基酸保留时间之和,范德华体积以及辛醇-水分配系数对多肽保留时间进行预测,采用KPLS进行建模预报,经过交叉验证得到很好的预测结果。更多还原

【Abstract】 Quantitative structure retention relationships(QSRR)were derived allowing prediction retention of peptides.To quantita- tively characterize the structure of a peptide,the following descriptors are employed:logarithm of the sum of retention times of the ami- no acids composing the peptide,logarithm of Van der Waals volume of the peptide,and logarithm of its calculated n-octanol-water par- tition coefficient,clog P.KPLS is used to nonlinear modeling and LOO is used to verify the results.The predicted gradient retention times are in excellent agreement with the experimental data.更多还原