What is the difference between Glide core RMSD as set up in the Core tab, and the feature "Compute RMSD to input ligand geometries" in the Output tab? Which should I use?

The "Compute RMSD to input ligand geometries" feature calculates the heavy-atom RMSD of each docked pose to its corresponding input structure. Obviously, the RMSDs will be meaningful only if the input ligand structures are already positioned in the frame of the receptor active site. The RMSD calculations on the Core tab compute the RMSDs of each docked pose against a single reference structure, using a subset of atoms defined (via SMARTS) by the user (i.e., the 'core'). Again, the RMSDs are computed 'in-place' (i.e., they include orientation/translation differences), so the reference molecule must be positioned in the frame of the receptor in order for the results to be meaningful. If you are docking ligands with a common core and want to compare how the cores of those ligands docked compared with a reference structure (e.g., the native co-crystallized ligand), you should use the Core tab RMSD calculations. Whichever RMSD method you use, the computed RMSDs will be stored as properties of the output poses.

For the special case in which you're redocking the native ligand and you want to compare the RMSD of the docked pose against the input structure, you can use either the Core tab (with the "All heavy atoms" option) or "Compute RMSD to input ligand geometries" on the Output tab. If you want to compare the docked pose to the original crystal structure (i.e., before the preparation procedure has modified any heavy-atom coordinates), you should use the Core tab. However, please make sure that the reference molecule (e.g., the crystal pose) is all-atom with correct bond orders, so the SMARTS pattern generated for the reference core will match that of the prepared, docked structure; the ligand structure generated by the Protein Preparation Wizard just after the Preprocess step should be suitable as the reference.

Back to Search Results