How to set selection string for superposition for RMSD for receptor alone, ligand alone and ligand with respect to the original binding site?

[HoaSoup]

Hello,

I need to do RMSD analysis to tell:

• If the receptor (protein, chain A) has converged to a certain conformation.
• If the ligand (the smaller protein, small molecule or peptide, chain B) has converged to a certain conformation.
• If the ligand has drifted away from the original binding site.

So, my question is: "How to set selection string for superposition for RMSD for receptor alone, ligand alone and ligand with respect to the original binding site?" I mean superposition as the step that superposes a selection of atoms before RMSD calculation takes place. For examples:

• Receptor alone: chainID A and backbone.
• Ligand alone (protein/peptide): chainID B and backbone.
• Ligand alone: chainID B and not name H*
• If the ligand has drifted away from the original binding site: chainID A and backbone.
  Are the examples above the correct selection strings for my purpose?

Best regards,
Hoa

[orbeckst]

You need to distinguish between what you're superimposing on to create a common frame of reference and what you're computing the RMSD of. The RMSD class will superimpose atomgroup on reference (modified by select) and then calculate the RMSD for atomgroup and and of the groups in groupselection.

I'd start by prototyping selections/atomgroups interactively and validate that it contains what you need. For example, for the receptor alone, your

ca = u.select_atoms("chainID A and backbone")

looks sensible.

If you want to see how the ligand moves relative to the protein, superimpose on the protein and list the ligand heavy atoms in groupselections.

If you want to understand if the ligand changes conformation itself, superimpose on the ligand and calculate the RMSD of the ligand to itself.
seems sensible.

[HoaSoup]

Thank you, @orbeckst.