%0 Journal Article
%T QSAR Model for Prediction of some Non-Nucleoside Inhibitors of Dengue Virus Serotype 4 NS5 using GFA-MLR Approach
%J Journal of Chemistry Letters
%I Iranian Chemical Science and Technologies Association
%Z 2717-1892
%A Adawara, Samuel Ndaghiya
%A Shallangwa, Gideon Adamu
%A Mamza, Paul Andrew
%A Ibrahim, Abdulkadir
%D 2020
%\ 07/01/2020
%V 1
%N 2
%P 69-76
%! QSAR Model for Prediction of some Non-Nucleoside Inhibitors of Dengue Virus Serotype 4 NS5 using GFA-MLR Approach
%K B3LYP/631G**
%K Quantitative structure-activity relationships
%K Genetic function algorithm
%K Molecular Descriptors
%R 10.22034/jchemlett.2020.114258
%X B3LYP/631G** basis set of DFT quantum mechanical method was used to optimize the molecular geometry of some non-nucleoside inhibitors of dengue 4 virus. Molecular descriptors were mined from the optimized structure and used along with their experimental inhibitory activity (pIC50) as the database for the study. Genetic function algorithm and multiple linear regressions were used to build a robust quantitative structure-activity relationship model. The statistically satisfactory quality of the model as evidenced by its validation parameters: R2 = 0.971, R2adj = 0.961, cRp^2 = 0.809 Q2 = 0.944 and R2pred = 0.627. Thus, the model can be used to predict the activity of new chemicals within its applicability domain. The Average Broto-Moreau autocorrelation - lag 1 / weighted by mass, Centered Broto-Moreau autocorrelation - lag 2 / weighted by Sanderson electronegativities, Coefficient sum of the last eigenvector from Barysz matrix / weighted by van der Waals volumes, nhigh lowest polarizability weighted BCUTS and Fraction of sp3 carbons to sp2 carbons are the descriptors that influenced the anti-dengue activity of the studied compounds. The information obtained from the model in this work can be employed to optimize the anti-dengue activity of the compounds.
%U