Homology modeling has been applied to fill in the gap in experimental G protein-coupled receptors structure determination.However,achievement of G protein-coupled receptors homology models with ligand selectivity remains challenging due to structural diversity of G protein-coupled receptors.In this work,we propose a novel strategy by integrating pharmacophore and membrane molecular dynamics(MD)simulations to improve homology modeling of G protein-coupled receptors with ligand selectivity.To validate this integrated strategy,the A2A adenosine receptor(A2AAR),whose structures in both active and inactive states have been established,has been chosen as an example.We performed blind predictions of the activestate A2AAR structure based on the inactive-state structure and compared the perfor-mance of different refinement strategies.The blind prediction model combined with the integrated strategy identified ligand-receptor interactions and conformational changes of key structural elements related to the activation of A2AAR,including(i)the movements of intra-cellular ends of TM3 and TM5/TM6;(ii)the opening of ionic lock;(iii)the movements of binding site residues.The integrated strategy of pharmacophore with molecular dynamics simulations can aid in the optimization in the identification of side chain conformations in receptor models.This strategy can be further investigated in homology modeling and expand its applicability to other G protein-coupled receptor modeling,which should aid in the discovery of more effective and selective G protein-coupled receptor ligands.