Computational QSAR Analysis of New Thiazol Derivatives as Inhibitors

In order to support the development of new 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1) inhibitors based on the pseudothiohydantoin scaffold and perhaps provide innovative treatments for metabolic illnesses, this research attempts to create a predictive model. 56 derivatives of 2-aminothiazol-4(5h)-one, for which the 11β-HSD1 inhibitory action was previously described, were subjected to the Quantitative Structure–Action Relationship (QSAR) study. Dragon software was utilised to compute the molecular descriptors, and Gaussian software was utilised for geometry optimisation. Regression analysis between the top ten preselected descriptors and the activity of the examined analogues was conducted using an Artificial Neural Network (ANN) technique. Using a network architecture 10-11-1 and a Broyden–Fletcher–Goldfarb–Shanno learning algorithm, a predictive model was created. Cross-validation and y-randomisation techniques bolstered the model's dependability. With a determination coefficient (R2) of 0.9482, the model demonstrated great accuracy, and internal validation verified its validity with a cross-validated R2 (Q2) of 0.9944. The QSAR model was created using four sets of topological indices (GALVEZ, 2D autocorrelations, 2D matrix-based descriptors, and Burden eigenvalues) and three classes of 3D descriptors (GETAWAY, 3D-MoRSE, and RDF descriptors). Compounds with cyclohexyl and 2-(tetrahydro-2H-pyran-2-yl)methyl residues substituted at the amino group and different substituents at C-5 of the thiazole ring may be viable candidates for future chemical synthesis and biological evaluation, according to the developed model, which was used to predict the 11β-HSD1 inhibitory activity of four designed series of 2-aminothiazol-4(5h)-one derivatives.