Virtual screening strategy for identifying new small-molecule antagonists of integrin IIb3

Aim. To develop an optimal strategy for identifying new small-molecule antagonists of integrin IIb3 using virtual screening. Methods. Molecular modeling. Results. The 7TMZ-based model demonstrated high classification accuracy (receiver operating characteristic area under the curve (ROC AUC): 84.285) and pose reproducibility (root mean square deviation (RMSD): 0.326 Å). The 3T3M-based model demonstrated high pose reproducibility (RMSD of 0.218 Å for RUC2 and 0.254 Å for RUC1). Conclusions. Two virtual screening models were developed to identify integrin IIb3 antagonists that do not induce receptor unfolding. Preliminary evaluation suggests their strong potential in selecting active compounds.