Skip to main content
Fig. 2 | Journal of Cheminformatics

Fig. 2

From: EDock: blind protein–ligand docking by replica-exchange monte carlo simulation

Fig. 2

An illustration of the graph matching method in EDock. b The ligand structure (in sticks) is to be matched with the binding pocket, which is represented by a set of indexed crosses. b Three illustrative nodes are highlighted in circles, including node1 (C:10), node2 (CD2:6), and node3 (CE2:3). The intra-ligand distances (\(d_{C - CD2}^{ligand} = 4.90\), \(d_{C - CE2}^{ligand} = 6.18\), and \(d_{CE2 - CD2}^{ligand} = 1.40\) Å) are close to the intra-pocket distances (\(d_{10 - 6}^{pocket} = 4.90\), \(d_{10 - 3}^{pocket} = 6.63\), and \(d_{3 - 6}^{pocket} = 2.00\) Å), which results in three edges (dashed lines) being added between the nodes following Eq. (1). The three-edged nodes thus form a graph clique. Based on the atom–atom correspondence specified by the graph clique, the initial docking poses are generated by superimposing the ligand structure with the grid points of the binding pocket using the Kabsch RMSD rotation matrix [16]

Back to article page