From: John Stone (
Date: Tue Dec 06 2016 - 22:37:04 CST

  You can have VMD edit PDB files by making atom selections and using
"$sel writepdb" to generate a new PDB file containing only the parts
you want to keep. This is described in the User's Guide and tutorials.
That would work much better than trying to use Excel for such things,
which is likely doomed due to the PDB file format being very sensitive to
exact spacing of the various fields. If you can write out two matching
substructures from VMD then you could load them back in, compute alignments
matrix, and then apply them to the original structure(s) you're working with,
as an alternative to other approaches.

  John Stone

On Wed, Dec 07, 2016 at 12:03:04AM +0000, Wong Li Zhe wrote:
> Hi Ashar,
> Many thanks for your detailed explanation.
> Not sure why but I didn't get any transformation matrix from Superpose,
> hence, I couldn't view it using VMD.
> Anyway, I used another server (PROSMALS3D) to run alignment and etc as
> what has been suggested by you.
> Managed to edit the PDB file manually using Excel and removed those
> unwanted residues.
> Prior to superimpose both protein structures on VMD, I tried converting
> from Excel to PDB but somehow it doesn't work with all those funny
> alphabets present in the file.
> I wonder do you have any insights on this?
> I have tried online conversion and it doesn't work.
> Any thoughts on this?
> Thank you in advance.
> Best regards,
> Li Zhe
> ---------------------------------------------------------------------------
> From: Ashar Malik <>
> Sent: Thursday, 1 December, 2016 4:42:56 PM
> To: Wong Li Zhe
> Cc:
> Subject: Re: vmd-l: Superimpose
> Hi,
> If you have used superpose to compare your two structures, as an output
> Superpose gives a transformation matrix (4x4). You can load both
> structures into VMD. You can then load the transformation matrix into vmd.
> You can then move one structure by applying the transformation matrix on
> it. You will then be able to see the two aligned structures.
> If you don't want to view superpose's output you can calculate your own.
> To this -- do the following.
> Open structure 1 in VMD.
> Look for sections of structure that form an alpha helix. If you see a
> section in the display that makes an alpha helix press 1 on the keyboard
> and using your mouse click on one of the residues in the helix. When you
> click VMD will generate a label of the residue number. Note that residue
> number. Say that number is 77. Probably 2-3 residues after residue 77 and
> before residue 77 are also involved in the helix. Which means residue 73
> to 80 from structure 1 are in the helix. Note these numbers.
> Delete the structure and load structure 2 in VMD.
> Repeat the same exercise as you did with structure 1. If you know what
> should align across the two structure good, if you don't take note of an
> alpha helix forming set of residues like in the previous case. Say you
> identify residue 384. For equivalence purpose if you selected 7 residues
> in the previous case [73 to 80] select seven residues from this structure.
> So if residue 384 is of interest, select 380 to 387. Take a note of
> these.
> Now you have two selections. The residues in these selections i.e. 73 to
> 80 in structure 1 and 380 to 387 may or may not be the same. This means
> that you still need to find something common --- this is your alpha
> carbons or backbone atoms which would be common between the two
> selections.
> Follow the following steps.
> Close VMD
> Open vmd
> load structure 1 (across which you liked residues 73 to 80)
> load structure 2 (across which you liked 380 to 387)
> --------------------------------------------------------------------------------------
> use the following commands:
> set sel0 [atomselect 0 "resid 73 to 80 and name CA"]
> set sel0_all [atomselect 0 all]
> set sel1 [atomselect 1 "resid 380 to 387 and name CA"]
> set mx_ [measure fit $sel0 $sel1]
> $sel0_all move $mx
> --------------------------------------------------------------------------------------
> Now your VMD display window will show the structure 1's residue 73 to 80
> aligned with structure 2's residue 380 to 387.
> This way you can choose what ever par from structure 1 to align with
> whatever part from structure 2.
> If you really want to do it your way and edit the PDB so that the two
> structures are similar size-- you can do it manually delete lines for
> residue 830 till 1304 or do this :
> close VMD if its open.
> Open VMD
> Load the larger structure into VMD:
> use the following command:
> set sel0 [atomselect 0 "resid 1 to 829"]
> $sel0 writepdb reduced_structure.pdb
> The above will create a new pdb file called reduced_structure.pdb which
> will only have residues 1 to 829 of the larger structure. Assuming there
> are no missing residues across the two structures, they should now be
> compareable (as long as you stick to the backbone atoms like entire
> backbone or CA etc). Including the side chains would still result in a
> different number of atoms. However this is a bad way to do it. You should
> know before hand which part to align, if you don't know you should use
> programs like Superpose to decide how to do this. Simply truncating a
> portion of the structure to fit your need is not the correct way. As a
> result of this truncation you may end up with a very poor match.
> Summary:
> If you don't know which parts align - use Superpose to align and use the
> transformation matrix in its output to move 1 structure onto the other.
> If you know which sections across the two structures align, use the method
> I stated above.
> If you want to cutoff a part of protein - do it by proper consideration
> instead of just editing a structure.
> Hope this helps.
> Best,
> /A
> p.s. when I say run the commands - you have to use the command line window
> (in windows this would be a dos like screen in the background), in
> linux/mac it would be the shell. Hit enter to get the prompt and type
> away. If you are not comfortable with that - goto extensions menu in VMD
> main and select TKconsole and enter commands there.
> On Thu, Dec 1, 2016 at 6:53 PM, Wong Li Zhe
> <[1]> wrote:
> Dear VMD users,
> I would like to superimpose 2 protein structures.
> It is however, they are of different size, 1304 and 829 respectively.
> I wonder how can I edit the PDB file manually in order to view them in
> VMD?
> I have tired using SuperPose version 1.0 to superimpose the structure,
> but I would like to view it in VMD.
> Any suggestions on how I should edit the protein sizes?
> Thanks in advance!
> Best regards,
> Li Zhe
> An opportunity to learn about our Foundation in Science (FiS) programme.
> Join many exciting activities for the day.
> Start your future in healthcare at our FiS Info Day!
> More at [2]
> --
> Best,
> /A
> An opportunity to learn about our Foundation in Science (FiS) programme.
> Join many exciting activities for the day.
> Start your future in healthcare at our FiS Info Day!
> More at
> References
> Visible links
> 1.
> 2.

NIH Center for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801           Phone: 217-244-3349