RMSD issue

77 views
Skip to first unread message

Keshav Patil

unread,
Mar 3, 2020, 12:47:00 PM3/3/20
to PLUMED users
Hi all,

I am facing a problem with using RMSD. So basically the way RMSD in plumed works is 

RMSD REFERENCE=file.pdb TYPE=OPTIMAL

where the RMSD is computed by aligning the atoms mentioned in file.pdb and the instantaneous position of those atoms.

However,  I am particularly interested in computing and biasing the RMSD for two sub domains of my system. So what I require is I align the structure for all but the two sub domains containing a certain range of residues  of my protein system and then compute the RMSD for the two sub domains.

But the way plumed does is it asks for reference file in which 

either 
 I could include only those sub domain residues and compute their RMSD but the issue is that these residues would be aligned and hence RMSD would be affected.

or
I could include all but those sub domain residues so that they get aligned and not sub domain residues, but in this case the RMSD won't be computed for the sub domain residues.

Can anyone please suggest a workaround ? Please let me know if further explanation is required to understand the issue. Thank You.

I am attaching my plumed script: The RMSD of the sub domains residues is what I am using as my CV.

-Keshav Patil
University of Pennsylvania

UNITS LENGTH=A ENERGY=kcal/mol

MOLINFO STRUCTURE=egfr_active.pdb

WHOLEMOLECULES RESIDUES=all MOLTYPE=protein


#RMSD to inactive structure alpha C helix

RMSD REFERENCE=egfr_inactive_CA_alphaChelix.pdb TYPE=OPTIMAL LABEL=inactive_rmsd_alphaChelix


#RMSD to active structure alpha C helix

RMSD REFERENCE=egfr_active_CA_alphaChelix.pdb TYPE=OPTIMAL LABEL=active_rmsd_alphaChelix


#RMSD to inactive structure activation loop

RMSD REFERENCE=egfr_inactive_CA_aloop.pdb TYPE=OPTIMAL LABEL=inactive_rmsd_aloop


#RMSD to active structure activation loop

RMSD REFERENCE=egfr_active_CA_aloop.pdb TYPE=OPTIMAL LABEL=active_rmsd_aloop


METAD ...

ARG=active_rmsd_alphaChelix,active_rmsd_aloop

SIGMA=0.05,0.05

HEIGHT=0.6

PACE=500

LABEL=meta

TEMP=300

BIASFACTOR=30

GRID_MIN=-1,-1

GRID_MAX=10,18

WALKERS_N=4

WALKERS_ID=0

WALKERS_DIR=../

WALKERS_RSTRIDE=500

... METAD

FLUSH STRIDE=500

ENERGY LABEL=energy

PRINT ARG=* STRIDE=500 FILE=COLVARS

Eliane Briand

unread,
Mar 4, 2020, 3:26:57 AM3/4/20
to PLUMED users
Hello,

As seen in the documentation (https://www.plumed.org/doc-v2.6/user-doc/html/_r_m_s_d.html), there is two set of weights that can be specified in the reference PDB (in place of occupancy and B-factor): one for structure alignement and one for displacement computation (RMSD proper). So I would suggest setting all alignement weights to 1.0, and all displacement weight to 0.0, except for the residues you're interested in which would get weight 1.0.

Would that solve your problem ?

Best regards.

Keshav Patil

unread,
Mar 13, 2020, 2:37:35 PM3/13/20
to PLUMED users
Yes it does. Thank You.

-Keshav
Reply all
Reply to author
Forward
0 new messages