Three-dimensional quantitative structure activity relationship modeling of the rate of penetration of diverse compounds through the human skin by using GRIND descriptors

The control of permeation is vital for the topical application of lotions, creams, and ointments, and for the toxicological and danger assessment of materials from environmental and occupational hazards. The skin permeability of industrial and household chemicals plays an important role in various fields including toxicology and risk assessment of hazardous materials, transdermal delivery of drugs and the plan of cosmetic products [1]. In this work, a three-dimensional quantitative structure–activity relationship (3DQSAR) model [2] has been developed to predict skin permeation of a variety of 210 compounds through human skin. Molecular descriptors were computed using GRid Independent Descriptors (GRIND) approach [3]. After variable selection via genetic algorithm method, 100 selected descriptors were correlated with skin permeability constants by various methods, such as PLS regression, support vector machine (SVM), and so on. Therefore, compounds are separated in two classification: 150 molecules as set of training and other compounds as set of test. The best results were obtained by PLS regression with the correlation coefficient of R2= 0.61 for calibration set. This strategy led to a final QSAR model that presented Q2= 0.59 and R2 pred=0.64 . As a result, the hydrogen bonding donor and hydrogen bonding acceptor of investigated compounds greatly influences their ability to penetrate through human skin. Furthermore, it has been observed that permeability was enhanced by increasing hydrophobicity and decreasing molecular weight.