rmsd and align on different chain in protein-protein

[aspitaleri]

Hi all,
my aim is to align to a chain A and calculate rmsd to chain B of a set of pdb files with respect to a single pdb:

import MDAnalysis as mda
from MDAnalysis.analysis import align, rms
from natsort import natsorted
import numpy as np
import pandas as pd
import glob
import sys

# Read reference structure and sorted PDB files
pdbref = "reference.pdb"
pdb_files = glob.glob("top_*.pdb")
pdb_files_sorted = natsorted(pdb_files)

print("Loading reference and trajectory files...")
ref = mda.Universe(pdbref)  
u = mda.Universe(*pdb_files_sorted, all_coordinates=True)

# Select Chain A for alignment and Chain B for RMSD calculation

mobile_chainB = u.select_atoms("chainID B and backbone")
reference_chainB = ref.select_atoms("chainID B and backbone")

# **Align trajectory on Chain A (Cα atoms)**
print("Aligning trajectory on Chain A...")
aligner = align.AlignTraj(u, ref, select="chainID A and backbone", in_memory=True)
aligner.run()

# **Compute RMSD on Chain B** AFTER alignment (manual calculation)
rmsd_values = []
for ts in u.trajectory:
    rmsd_value = rms.rmsd(mobile_chainB.positions, reference_chainB.positions, center=False, superposition=False)
    rmsd_values.append([ts.frame, ts.time, rmsd_value])

# Save RMSD results
df = pd.DataFrame(rmsd_values, columns=["Frame", "Time", "RMSD_ChainB"])
df.to_csv("rmsd_chainB.csv", index=False)

is that correct approach?
Thanks

[orbeckst]

Assuming that all your pdb_files = glob.glob("top_*.pdb") are the same number of atoms then you can read them as a trajectory.

You can use Aligner and rmsd manually but most people would likely just use rms.RMSD (see also the User Guide example on RMSD.