QSRR modeling for diverse drugs using different feature selection methods coupled with linear and nonlinear regressions

摘要: Abstract A Quantitative Structure-Retention Relationship (QSRR) is proposed to estimate the chromatographic retention of 83 diverse drugs on a Unisphere poly butadiene (PBD) column, using isocratic elutions at pH 11.7. Previous work has generated QSRR models for them Classification And Regression Trees (CART). In this work, Ant Colony Optimization used as feature selection method find best molecular descriptors from large pool. addition, several other methods have been applied, such Genetic Algorithms, Stepwise and Relief method, not only evaluate but also investigate its ability important in QSRR. Multiple Linear (MLR) Support Vector Machines (SVMs) were applied linear nonlinear regression methods, respectively, giving excellent correlation between experimental, i.e. extrapolated mobile phase consisting pure water, predicted logarithms factors (log kw). The overall model was SVM one built selected by ACO.