Linear and non-linear QSAR models on Paclitaxel derivatives as Anticancer drugs

This research presents a quantitative structure-activity relationship (QSAR) of IC50 values of Paclitaxel derivatives. Twenty five different Paclitaxel anticancer derivatives were selected as asample set and the geometry of the complexes were optimized using Gaussian 03W. The activity of the 25 different Paclitaxel derivatives was estimated by means of multiple linear regression (MLR), artificial neural network (ANN), and genetic algorithm (GA) techniques.These models were also utilized to select the most efficient subsets of descriptors in a crossvalidation procedure for non-linear -log(IC50) prediction. The results obtained using the GAANNwere compared with those obtained using MLR- PLS1, MLR-ANN ,GA-MLR and GAANNmethodes. A high predictive ability was observed for the MLR-PLS1, MLR-ANN ,GAMLR GA-ANN models, with root mean sum square errors (RMSE) of 0.362613, 0.0478, 0.1304, 0.0720 respectively (N=25). The results obtained using the GA-ANN method indicated that the activity of the derivatives of Paclitaxel depends on different parameters such as GATS7v ,Mor28m, Mor32p,ATS8e,ESPm03u,RDF035m,Mor06p and MLOGP descriptors. In summary, acomparison of the quality of ANN with different MLR methods showed that ANN has a better Predictive ability.

کلیدواژه ها:

Paclitaxel ، Antitumor drugs ، QSAR ، Gentic Algorithm