QSAR study of 1-(3, 3-diphenylpropyl)-piperidinyl amides and ureas using genetic algorithms and artificial neural networks. (2016)

• Record Type: Journal Article
• Authors: Labjar, Houda; Kissi, Mohamed; Mouhibi, Rokaya; Khadir, Omar; Chaair, Hassan; Zahouily, Mohamed
• Is Part Of: International journal of bioinformatics research and applications. Volume 12:Number 2(2016)
• Page Start: 116
• Page End: 128
• Publication Date: 2016
• Subjects: QSAR -- quantitative structure activity relationships -- genetic algorithms -- artificial neural networks -- ANNs -- CCR5 -- multiple linear regression -- 1-(3, 3-diphenylpropyl)-piperidinyl amides -- ureas -- molecular orbital -- molecular weight -- molecular volume -- molar refractivity -- LogP parameters -- human immunodeficiency virus -- HIV -- acquired immune-deficiency syndrome -- AIDS -- bioinformatics; Bioinformatics -- Periodicals; 570.285
• Languages: English
• ISSNs: 1744-5485

Abstract

Quantitative Structure Activity Relationships (QSAR) were studied for a series of 54 1-(3, 3-diphenylpropyl)-piperidinyl amides and ureas derivatives by means of Multiple Linear Regression (MLR), Genetic Algorithm (GA) and Artificial Neural Network (ANN) techniques. The values of pIC50 (dose of compound required to reduce the proliferation of normal uninfected cells by 50%) of the studied compounds were correlated with the descriptors or variables encoding the chemical structures. An approach that combines GA and MLR (GA-MLR) was used to select the pertinent descriptors to explain the activity pIC50 . The descriptors revealed by GA-MLR were used to characterise the non-linear aspect in the activity parameter. The results obtained from this study indicate that the activity pIC50 is strongly dependent on the highest occupied molecular orbital, molecular weight, molecular volume, molar refractivity and LogP parameters.