Skip to main content
Fig. 9 | Journal of Cheminformatics

Fig. 9

From: Binding mode information improves fragment docking

Fig. 9

Variation of GRIM performance in pose prediction. The proportion of correct poses is based on the RMSD between the native and the docked poses of the fragment, considering five threshold values. Poses were selected by comparison to reference drug-like ligands and fragments (DF). a Filtering of low quality poses. Pose selection using the reference dataset before filtering with EDIA score (All, 832 docked fragments, 2082 reference molecules) and after filtering with EDIA score (Filt, 586 docked fragments, 1529 reference molecules). Filt is the dataset described in this study. b Native docking (Native) is compared to cross-docking using all the structures of the protein site (Cross) or only its representative structure (Ref)

Back to article page