A set of nineteen Pyrazinecarboxamides derivatives with herbicidal activity was subjected to the two dimensional quantitative structure activity relationships studies. This work was conducted using the principal component analysis (PCA) method, the multiple linear regression method (MLR), the multiple non-linear regressions (MNLR) and the artificial neural network (ANN). The predicted results of various study compounds afford reliable prediction of IC50 with respect to experimental data. Density functional theory (DFT) calculations have been carried out in order to get insights into the structure, chemical reactivity and property information for the series of study compounds. This study shows that the PCA, MLR and MNLR have served also to predict activities, but when compared with the results given by the ANN (R2= 0.994) , we realized that the predictions fulfilled by this latter were more effective as indicated by the value of cross validated squared correlation coefficient (R2 CV = 0.998). Thus, this validated model brings important structural insight to aid the design of novel herbicidal agents.
