Kết hợp mô hình phân tử và protein docking trong thiết kế chất ức chế yếu tố đông máu hoạt hóa Xa mang khung isosteviol

The present paper describes in silico design of novel oxime ether isosteviol derivatives as well as molecular modeling approach based on structure-activity relationship analysis (QSAR) and docking simulation for searching novel isosteviol-based compounds as potential FXa inhibitors. Method: Automated Network Search (ANS) algorithm was employed for constructing QSAR models to predict the FXa inhibitory activities. New compounds bearing isosteviol moieties were designed based on the structure-activity relationships obtained from QSAR models. Docking simulations using Autodock Vina program were applied for predicting the binding ability of the most active compounds towards FXa. Results: A QSAR model with R2 training of 0.93, and R2 external validation of 0.84 was obtained based on a multilayer perceptron (MLP) 5-8-1 network and a database of 77 isosteviols. 16 new derivatives were designed, and their binding energies ranged between - 6.9 to -9.1 kcal/mol. Among them, 5 isosteviols showed the highest activity against FXa. Conclusions: Analysis showed that the most promising derivatives contain heterocyclic aromatic 􀀕ve-membered moieties with substituents containing chlorine or 􀂣uorine atoms. It is anticipated that the 􀀕ndings reported in present 27 work may provide useful information for designing e􀂤ective FXa inhibitors as anticoagulant agents.