Optimized Ensembled Predictive Model for Drug Toxicity

Healthcare is one of the most important concerns for living beings. Prediction of the toxicity of a drug is a great challenge over the years. It is quite an expensive and complex process. Traditional approaches are laborious as well as time-consuming. The era of computational intelligence has started and gives new insights into drug toxicity prediction. The quantitative structure-activity relationship has accomplished significant advancements in the field of toxicity prediction. Nine machine learning algorithms are considered such as Gaussian Process, Linear Regression, Artificial Neural Network, SMO, Kstar, Bagging, Decision Tree, Random Forest, and Random Tree to predict the toxicity of a drug. In the study, we developed an optimized regression model (Optimized KRF) by ensembling Kstar and Random Forest algorithm. For the mentioned machine learning models, evaluation parameters are assessed. The 10-fold cross-validation is applied to validate the model. The optimized model gave a coefficient of correlation, coefficient of determination, mean absolute error, root mean squared error, and accuracy of 0.9, 0.81, 0.23, 0.3, and 77% respectively. Further, the Saw score is calculated in two aspects as W-Saw score and the L-Saw score. The W-Saw score for the optimized ensembled model is 0.83 which is the maximum and L-Saw score is 0.27 which is the lowest in comparison to other classifiers. Saw score provides the strength to an ensemble model. These parameters indicate that the optimized ensembled model is more reliable and made predictions that were more accurate than earlier models. As a result, this model could be efficiently utilized to forecast toxicity.