| Abstract [eng] |
Comparative and computational research on binding constants Kd and structures of 40 CA II inhibitors, i.e. benzensulfonamide derivatives with an attached pyrimidine ring, was carried out. Docking, LIE, metadynamics, and QSAR methods were used for the comparison. The computed Kd values were compared with the experimental data. This enabled the evaluation of advantages and disadvantages of the methods used when applied for the CA II receptor and its inhibitors. The best results were obtained with QSAR (coefficient R2 between the experimental and predicted pKd values varied from 0.83 to 0.89). Possible reasons for the observed results were explored. Moreover, a new improvement for the LIE method was suggested. With the proposed energy estimation equation, LIE results noticeably improved (R2 between the experimental and predicted ΔGbind values improved from 0.24 to 0.50). Using the antiproliferative activity data of α-branched α,β-unsaturated ketones against NB4, MCF–7, and A549 cancer cell lines, several QSAR models using 5 descriptors were developed. They can be used to predict and explain the antiproliferative activities of compounds belonging to this class. A probability to find novel inhibitors with a novel inhibitor search strategy by employing the Tanimoto similarity index between compound candidates and metabolites was estimated. This strategy led to a discovery of two compounds inhibiting cancer cell growth, using hitherto unknown modes of action. |
