calculating rmsd for atoms of each residue

[Divya Singhal]

Hi..

I am a beginner to use MDAnalysis and trying to calculate the rmsd values of atoms for each residue. I am using following code to calculate this:

u = mda.Universe('structure.pdb', 'rmsd.dcd') 

ref=mda.Universe('referance.pdb')

per_res_rmsd=np.empty((800,94),float)

for r in range (94):

    atom=(u.residues[r]).atoms

    atom_ref=(ref.residues[r]).atoms

    R = rms.RMSD(atom,  # universe to align

            atom_ref,  # reference universe or atomgroup

             select='resid 1-94')  # group to superimpose and calculate

             #groupselections=[GRP])  # groups for RMSD

             #ref_frame=0)  # frame index of the reference

    R.run()

    per_res_rmsd[:,r]=R.rmsd[:,2]

But i am not sure whether I am doing it right or not. Please help

    

[Oliver Beckstein]

Hi Divya Singhal,

Do you want the RMSD for each residue and frame after (a) superposition of each residue on itself in the reference, or (b) superposition of all residues on all residues in the reference? (a) would measure the internal configurationial flexibility, (b) how much each residue contributes to the overal RMSD.  Depending on (a) or (b), different approaches apply.

I am asking because your code snippet contains "select='resid 1-94')  # group to superimpose and calculate” which points towards (b) being intended but "atom_ref,  # reference universe or atomgroup” being "atom_ref=(ref.residues[r]).atoms” actually does (a). 

For (b) you could use "groupselections=[f”resid {resid}” for resid in range(1, 95)]”. For (a) you have to something like what you’re already doing.

Oliver

--

Oliver Beckstein (he/his/him)

email: orbe...@mdanalysis.org
twitter: @orbeckst

MDAnalysis – a NumFOCUS fiscally sponsored project

https://www.mdanalysis.org/