LAUSR

Actin-binding kinases are enzymes involved in the regulation of actin polymerization and microtubule disassembly. The inhibition of actin-binding kinases can be used in the treatment of glaucoma, since it relaxes the tissue, increases outflow facility, and reduces intraocular pressure. This study presents quantitative structure–activity relationship (QSAR) modeling for a group of 59 chemical compounds, actin-binding kinase inhibitors. The modeling was based on the Monte Carlo optimization with molecular descriptors based on the simplified molecular-input line-entry system notation and local invariants of the molecular graph, as well as on 3D field-based methods. Conformation independent QSAR models were developed for three random splits, whereas the 3D QSAR model was developed for one random split into the training and test set. Regarding the statistical quality of the developed models, including robustness and predictability, it was tested using numerous statistical approaches and the obtained results proved good. There was an excellent correlation between the results from the 3D QSAR and conformation independent models. In addition, a new statistical metric – the index of ideality of correlation was applied for the final assessment of the model, and the results were good. Molecular fragments responsible for the increases and decreases of a studied activity were defined and used for the computer-aided design of new compounds as potential actin-binding kinase inhibitors. The final assessment of the developed QSAR model and designed inhibitors was done using molecular docking studies, which were in excellent correlation with the results from QSAR modeling.