The mammalian olfactory system comprises a large family of G protein-coupled receptors (GPCRs) to detect and discriminate numerous volatile ligands. More than 350 human genes encode functional olfactory receptors (ORs) [1] that belong to the class A (rhodopsin-like) GPCR family. [2] Owing to difficulties with functional OR expression in heterologous systems, [3,4] only a few human ORs have been characterized to date. Most deorphanized ORs, that is, ORs with a known ligand spectrum, detect multiple chemically similar odorants, [5] and hypervariable residues in the seven transmembrane (7TM) helices (I-VII) have been postulated to form the basis for ligand specificity.[6] A prerequisite for understanding olfactory receptor selectivity is information on the spatial properties of the ligand-binding niche. Different classes of approaches have been employed for such an assessment: [7] Ligand-based approaches, such as pharmacophore modeling or quantitative structure-activity relationship (QSAR), can give valuable models of the ligand structure, which is required for discriminating activating and inactive ligands, and information on the form of the binding pocket. [8][9][10][11] Receptor-based approaches such as homology modeling create a model of the protein and the binding site explicitly, [12][13][14][15][16][17][18] and from this give information on ligand binding. Both techniques can be combined together. [19][20][21][22] For such receptor-based and mixed approaches, the X-ray structures of seven GPCRs have been solved to date, [23][24][25][26][27][28][29][30][31][32] but none for ORs. Previous studies have used static structural models of different ORs based on a rhodopsin [12][13][14][15][16][17][18][19][20][21] and a b 2 -adrenergic receptor (B2AR) [33] template. However, most odorants are highly flexible, so assessment of the ligand/protein dynamics might be of crucial importance in understanding ligand recognition by ORs. To better understand receptor activation, we thus searched for a dynamic ligand-protein interaction pattern instead of analyzing ligand-binding in static models. Therefore, in difference to other flexible GPCR ligand pocket analysis approaches, [34][35][36][37] we use the predictive power of protein/ligand complex molecular dynamics (MD) simulations [38][39][40] to gain insight into the protein-odorant dynamics necessary for receptor activation.We developed a dynamic model of the functionally wellcharacterized human olfactory receptor hOR2AG1.[41] We used an X-ray structure of bovine rhodopsin with 2.2 resolution [24] as starting structure for dynamic homology modeling of hOR2AG1, since both receptors belong to the class A GPCRs, and both harbor hydrophobic ligands. The performance of this approach was previously tested by homology modeling of the B2AR ligand-binding niche based on the rhodopsin template (see Supporting Information, Section 1 a).[40] Although the overall sequence identity among class A GPCRs is relatively low, this can be compensated for by careful incorporation of experimental...