superimposing two structures based on a specific selection
Chimera 1.8 on Windows I have been trying to align (superimpose) models based on a specific conserved region. Unfortunately Chimera doesn't seem to do what the documentation claims it does. I run (example): match #0:5-15@CA,20-30@CA #1: 5-15@CA,20-30@CA and get an RMSD calculation (more on this below), but no movement of the model occurs. I specify 'move true show true' as options. Still no movement but I have a transformation matrix. Unfortunately this is not in a form that can be understood by Transform Coordinates. How do I persuade my structures to superimpose on the specific atoms? The RMSD calculation appears to be broken. If I take the command rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,20-30@CA then I will get a number If I then run rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,21-31@CA I get exactly the same number. In fact, if I specify atom sets of different sizes (as long as it is not the first block) then I continue to get exactly the same number rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,20-35@CA This should fail but doesn't Have I completely misunderstood what Chimera should be doing here? ..d The University of Dundee is a registered Scottish Charity, No: SC015096
Hi David, I tried exactly your command in Chimera 1.8 with two randomly chosen model (3pfj, 2gbp) and match worked as expected (after removing the space after "#1:" which gives an error). So to know why your case does not work we'd need the actual PDB models you are using. But without those here is my suspicion. The match doesn't move the first model because it is already aligned. You could test this by moving one model relative to the other, then try the command again (use the Active buttons in Model Panel to freeze one model, then move the other with the mouse). Also the specifier "#0:5-15@CA,20-30@CA" has the wrong syntax. It should be #0:5-15@CA:20-30@CA" (with the comma replaced by a colon). The first specifier will only select the 11 CA atoms 5-15 because with the comma it says take atoms named CA and atoms named "20-30" and atoms named CA. There are no atoms named "20-30". This explains why your rmsd values don't depend on the 20-30 or 21-31 or 20-35. Here is some documentation on atom specifiers http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/frameatom_spec.html Note that the 7th example is the same kind of specifier you were attempt with two residue ranges. Tom On Feb 10, 2014, at 3:08 AM, David Martin <d.m.a.martin@dundee.ac.uk> wrote:
Chimera 1.8 on Windows
I have been trying to align (superimpose) models based on a specific conserved region. Unfortunately Chimera doesn’t seem to do what the documentation claims it does.
I run (example):
match #0:5-15@CA,20-30@CA #1: 5-15@CA,20-30@CA and get an RMSD calculation (more on this below), but no movement of the model occurs. I specify ‘move true show true’ as options. Still no movement but I have a transformation matrix. Unfortunately this is not in a form that can be understood by Transform Coordinates.
How do I persuade my structures to superimpose on the specific atoms?
The RMSD calculation appears to be broken. If I take the command
rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,20-30@CA then I will get a number
If I then run rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,21-31@CA
I get exactly the same number. In fact, if I specify atom sets of different sizes (as long as it is not the first block) then I continue to get exactly the same number
rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,20-35@CA
This should fail but doesn’t
Have I completely misunderstood what Chimera should be doing here?
..d
The University of Dundee is a registered Scottish Charity, No: SC015096 _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
Hi David, A way to see if your atom specifier is doing what you expect is to try it with the "select" select command select #0:5-15@CA:20-30@CA Then see if the correct atoms are selected in the graphics, or look at the status line and see if the correct number of atoms were selected. This is the technique I used to see that your atom specifier was wrong. Tom On Feb 10, 2014, at 10:28 AM, Tom Goddard wrote:
Hi David,
I tried exactly your command in Chimera 1.8 with two randomly chosen model (3pfj, 2gbp) and match worked as expected (after removing the space after "#1:" which gives an error). So to know why your case does not work we'd need the actual PDB models you are using. But without those here is my suspicion. The match doesn't move the first model because it is already aligned. You could test this by moving one model relative to the other, then try the command again (use the Active buttons in Model Panel to freeze one model, then move the other with the mouse).
Also the specifier "#0:5-15@CA,20-30@CA" has the wrong syntax. It should be #0:5-15@CA:20-30@CA" (with the comma replaced by a colon). The first specifier will only select the 11 CA atoms 5-15 because with the comma it says take atoms named CA and atoms named "20-30" and atoms named CA. There are no atoms named "20-30". This explains why your rmsd values don't depend on the 20-30 or 21-31 or 20-35.
Here is some documentation on atom specifiers
http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/frameatom_spec.html
Note that the 7th example is the same kind of specifier you were attempt with two residue ranges.
Tom
On Feb 10, 2014, at 3:08 AM, David Martin wrote:
Chimera 1.8 on Windows
I have been trying to align (superimpose) models based on a specific conserved region. Unfortunately Chimera doesn’t seem to do what the documentation claims it does.
I run (example):
match #0:5-15@CA,20-30@CA #1: 5-15@CA,20-30@CA and get an RMSD calculation (more on this below), but no movement of the model occurs. I specify ‘move true show true’ as options. Still no movement but I have a transformation matrix. Unfortunately this is not in a form that can be understood by Transform Coordinates.
How do I persuade my structures to superimpose on the specific atoms?
The RMSD calculation appears to be broken. If I take the command
rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,20-30@CA then I will get a number
If I then run rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,21-31@CA
I get exactly the same number. In fact, if I specify atom sets of different sizes (as long as it is not the first block) then I continue to get exactly the same number
rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,20-35@CA
This should fail but doesn’t
Have I completely misunderstood what Chimera should be doing here?
..d
The University of Dundee is a registered Scottish Charity, No: SC015096 _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
Thankyou - it was my misreading of the atom specification pages that has lead to this issue. Many thanks for your help. ..d From: Tom Goddard [mailto:goddard@sonic.net] Sent: 10 February 2014 18:29 To: David Martin Cc: 'chimera-users@cgl.ucsf.edu' Subject: Re: [Chimera-users] superimposing two structures based on a specific selection Hi David, I tried exactly your command in Chimera 1.8 with two randomly chosen model (3pfj, 2gbp) and match worked as expected (after removing the space after "#1:" which gives an error). So to know why your case does not work we'd need the actual PDB models you are using. But without those here is my suspicion. The match doesn't move the first model because it is already aligned. You could test this by moving one model relative to the other, then try the command again (use the Active buttons in Model Panel to freeze one model, then move the other with the mouse). Also the specifier "#0:5-15@CA,20-30@CA" has the wrong syntax. It should be #0:5-15@CA:20-30@CA" (with the comma replaced by a colon). The first specifier will only select the 11 CA atoms 5-15 because with the comma it says take atoms named CA and atoms named "20-30" and atoms named CA. There are no atoms named "20-30". This explains why your rmsd values don't depend on the 20-30 or 21-31 or 20-35. Here is some documentation on atom specifiers http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/frameatom_spec.html Note that the 7th example is the same kind of specifier you were attempt with two residue ranges. Tom On Feb 10, 2014, at 3:08 AM, David Martin <d.m.a.martin@dundee.ac.uk<mailto:d.m.a.martin@dundee.ac.uk>> wrote: Chimera 1.8 on Windows I have been trying to align (superimpose) models based on a specific conserved region. Unfortunately Chimera doesn't seem to do what the documentation claims it does. I run (example): match #0:5-15@CA,20-30@CA #1: 5-15@CA,20-30@CA and get an RMSD calculation (more on this below), but no movement of the model occurs. I specify 'move true show true' as options. Still no movement but I have a transformation matrix. Unfortunately this is not in a form that can be understood by Transform Coordinates. How do I persuade my structures to superimpose on the specific atoms? The RMSD calculation appears to be broken. If I take the command rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,20-30@CA then I will get a number If I then run rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,21-31@CA I get exactly the same number. In fact, if I specify atom sets of different sizes (as long as it is not the first block) then I continue to get exactly the same number rmsd #0:5-15@CA,20-30@CA #1: 5-15@CA,20-35@CA This should fail but doesn't Have I completely misunderstood what Chimera should be doing here? ..d The University of Dundee is a registered Scottish Charity, No: SC015096 _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu<mailto:Chimera-users@cgl.ucsf.edu> http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users The University of Dundee is a registered Scottish Charity, No: SC015096
participants (2)
-
David Martin -
Tom Goddard