Modeling the QSAR of ACE Inhibitory Peptides Based on MLR
Xiao Ni Guan · Journal of Analytical Science · 2012
The structures of ACE inhibitory dipeptides,tripeptides and tetrapeptides with competitive type were characterized with amino acid descriptors SVHEHS.Furthermore,the multiple linear regression(MLR) models between the structure and the activities were built.The correlative coefficient(R2),the crossvalidation correlative coefficient(Q2LOO),root mean square error(RMSE),and external validation correlative coefficient(Q2ext) were 0.851,0.781,0.327,0.792 for dipeptides model;0.805,0.717,0.339,0.817 for tripeptides model;and 0.792,0.553,0.393,0.630 for tetrapeptides model,respectively.The results showed that the MLR models constructed by this descriptor had good fitting and predictive abilities,and could be used for quantitative structure activity relationship(QSAR) research of angiotensinⅠ-converting enzyme-inhibitory peptides.