Informatics Publishing Limited

Fatiha Mebarki ¹, Khadidja Amirat ¹, Salima Ali Mokhnache ¹, Djelloul Messadi ¹

Affiliations
1 Laboratoire de Securite Environnementale et Alimentaire, Universite Badji Mokhtar Annaba, B.P. 12, 23000 Annaba, DZ

Abstract

The gas chromatographic retention indices for (89 pyrazines of test and 25 of validation) on O V-101 and Carbowax -20M are successfuty modeled with the ald of a computer and the Software system. Structural descriptors are calculated and multiple linear regression analysis are used to generate model equations relating structural features to observed retention characteristics then was treated with two methods . The detection of influential observations for the standard least squares regression model is a problem which has been extensively studied. LAD regression diagnostics offers alternative dicapproaches whose main feature is the robustness. Here a nonparametric method for detecting influential observations is presented and compared with other classical diagnostics methods. Comparisons are between models generated for the two stationary was carried out with two methods, and descriptors that may encode differences in solute interactions with stationary phases of differing polarity are discussed and validated results in the state approached by the tests statistics: Test of Anderson-Darling, shapiro-wilk, Agostino, Jarque-Bera and two graphic test: Q-Q-plot of the residues, Histogram of frequency of the error and the confidence interval thanks to the concept of robustness to check if the distribution of the errors is really approximate.

Keywords

LAD Regression, Robustness, Outliers, Leverage Points, Tests Statistics.

Full Text

   

Fatiha Mebarki ¹, Souhaila Meneceur ², Abderrhmane Bouafia ³

Affiliations
1 1Department of Material Sciences, Faculty of Science and Technology, Amine Elokkal Elhadj Moussa Eg Akhamouk University – Tamanrasset 11000, Algeria., IN
2 Process Engineering Department, Faculty of Technology, Hamma Lakhdar University-El oued 3900, Algeria., IN
3 Process Engineering Department, Faculty of Technology, Hamma Lakhdar University-El oued 3900, Algeria ., IN

Abstract

The importance of Chemometrics Methods in Modeling (in QSAR analysis) of the mathematical model’s study of large datasets of molecules with huge numbers of physicochemical and structural parameters quantitative structure-Toxicity relationship (QSTR) are mainly based on multiple regression analysis in QSAR analysis The study of Least Square in deriving QSTR models for datasets of Quantitative Structure-Toxicity relationship on Log k ow (Octanol-water partition coefficient) for 16 Hydrocarbons compounds has been using the software Hyperchem 6.3 for computing descriptors and MINITAB 16 for data modeling. A three -descriptors model [two electronics molecules’ descriptors (QSER descriptor), HOMO (is Highest occupied molecular orbital) and LUMO (Lowest unoccupied molecular orbital), one QSAR descriptor 𝐸𝐻 (Hydration Energy) by Least Squares with correlation coefficient r=0.868, S=0.635, R 2 = 75.4%, R 2 ajd=73.7% and Durbin-Watson statistic =1.85277 and graphical analysis by diagram of goodness of fit and line plot. The results statistical of new model after removing the aberrant compounds (Toxicity compounds) shows high Coefficient of correlation r=0.9581, S=0.4316, determination coefficient R 2 =91. 8%, ajustemed R 2 ajd = 89.3%, Durbin-Watson statistic D=2.373, Three explanatory Variable model selected is robust and has good fitness. Two influential compounds detected and important the model and absence aberant compounds of the studied sample.

Keywords

Hydrocarbons (PAHs), toxicity 𝐿𝑜𝑔 𝑘𝑜𝑤,HOMO, LUMO,𝐸𝐻,Multiple linearregression, Aberant compounds.

Full Text