From: Jacob Keller
Date: 28 November 2011 22:22
Dear Crystallographers,
is there a ccp4 program--or otherwise--which can compute ca-ca
distances of corresponding residues between two superposed structures?
Jacob
----------
From: Bosch, Juergen
----------
From: Mario Sanches
Hi Jacob,
----------
From: Jacob Keller
Let me refine my question (sorry for my lack of clarity): is there a
program that will output the distances between the corresponding ca's
of a superposition on a residue-by-residue basis, and not just a
global RMSD value (doubtless these numbers are part of the
superposition algorithm itself)? I want to plot these values as a
function of residue number to show which parts of the structures
deviate more or less from each other.
Jacob
----------
From: Abhinav Kumar
CCP4's superpose does this.
It produces a table like this, but only for residues that are used in the alignment.
.-------------.------------.-------------.
| Query | Dist.(A) | Target |
|-------------+------------+-------------|
| + A:ARG 10 | |H+ A:GLN 223 |
| + A:ARG 11 | |H+ A:ASN 224 |
|H+ A:LYS 12 | <..6.19..> |H+ A:PRO 225 |
|H+ A:LYS 13 | <--4.03--> |H+ A:ASN 226 |
|H+ A:LYS 14 | <--3.73--> |H+ A:GLN 227 |
|H+ A:GLN 15 | <..4.05..> |H- A:LEU 228 |
|H+ A:LYS 16 | <..3.41..> |H- A:ILE 229 |
|H+ A:GLU 17 | <..1.75..> |H. A:SER 230 |
|H- A:ILE 18 | <++1.97++> |H- A:LEU 231 |
Thanks,
Abhinav
----------
From: Robert Immormino
Hi Jacob,
This sounds like what you get from ddmp (
http://www.csb.yale.edu/userguides/datamanip/ddmp/ddmp_descrip.html ).
Good Luck,
-bob
----------
From: Eric Pettersen
----------
From: Edward A. Berry
The program:
http://sb20.lbl.gov/berry/for/pdbdist2b.for
does this.
If you run it by the wrapper pdbd2b:
echo 'Find distances greater than threshold between corresponding atoms in 2 PDB files'
echo 'Usage: pdbd2b file1 file2 startres# [thresh]'
pdbdist2b <<eof
$1
$2
$3 10000
$3
$4 3.0
eof
then the format is:
pdbd2b file1 file2 startresiduenumber threshold
(set the threshold to -1 to print all distances)
then it is easy to redirect the output table to a textfile which can be
imported to scalc or excel for plotting.
It is kind of stupid: ignores chain letters, and once it finds
startresiduenumber in both files it proceeds without looking at numbers
or chains, so the atoms have to be in the same order in both files.
But if you have an insertion deletion you can just run twice, starting after the midmatch the second time.
N GLU A 13 N GLU A 13 0.0105
CA GLU A 13 CA GLU A 13 0.0092
CB GLU A 13 CB GLU A 13 0.0186
CG GLU A 13 CG GLU A 13 0.0369
CD GLU A 13 CD GLU A 13 0.0490
OE1 GLU A 13 OE1 GLU A 13 0.0623
OE2 GLU A 13 OE2 GLU A 13 0.0494
C GLU A 13 C GLU A 13 0.0231
O GLU A 13 O GLU A 13 0.0648
N THR A 14 N THR A 14 0.0156
CA THR A 14 CA THR A 14 0.0112
CB THR A 14 CB THR A 14 0.0160
OG1 THR A 14 OG1 THR A 14 0.0164
CG2 THR A 14 CG2 THR A 14 0.0273
C THR A 14 C THR A 14 0.0157
-
written for g77 but seems to compile with gfortran
I need to look in Dave's calendar prog to remind me how to get
command lines parameters in fortran - then can dispense with the wrapper.
----------
From: Tim Gruene
Dear Jacob,
Coot prints this information to the terminal, so you can start coot and
'tee' its output into a file.
Tim
- --
- --
----------
From: Jacob Keller
Dear Crystallographers,
Thanks very much for all who responded to my request. Below is a
compiled list of the various ways to skin this crystallographic cat!
Jacob
-Moleman ca plot distance-CCP4's superpose-distance matrix analysis,
e.g. http://www.csb.yale.edu/userguides/datamanip/ddmp/ddmp_descrip.html,
or other programs-Lsqkab, with the "deltas" card-If you use the
you want. OR If you already have the structures superimposed on your
RMSD header.-Coot prints this information to the terminal, so you can
start coot and 'tee' its output into a file.-Various fortran/homebrew
programs available from individuals-LGA server:
http://proteinmodel.org/AS2TS/LGA/lga.html
----------
From: Eleanor Dodson
Dont you get that from SSM and lsqkab (you have to turn on - list atom distances or some such in the GUI)
Eleanor
----------
From: Gerard DVD Kleywegt
You mean to produce something like this? http://xray.bmc.uu.se/usf/pics/distplot_1chr.gif
That can be done with LSQMAN - http://xray.bmc.uu.se/usf/lsqman_man.html#S65
Of course, there are lots of other interesting metrics and plots to assess structural differences between two models - http://xray.bmc.uu.se/usf/lsqman_man.html#H17
One of my favourite ones is the CD plot, for comparing multiple models, which looks a bit like a gel: http://xray.bmc.uu.se/usf/pics/cdplot_1ldn.gif
For other options, see http://xray.bmc.uu.se/usf/lsqman_man.html#H19
--DVD
******************************************************************
Gerard J. Kleywegt
Date: 28 November 2011 22:22
Dear Crystallographers,
is there a ccp4 program--or otherwise--which can compute ca-ca
distances of corresponding residues between two superposed structures?
Jacob
----------
From: Bosch, Juergen
I'd say you have to look into the USF toolbox e.g. Moleman2
You want to calculate rmsd's there's an option for that.
Jürgen
----------
From: Mario Sanches
Hi Jacob,
I used Moleman to do that once (http://xray.bmc.uu.se/usf/moleman2_man.html)
The Distance LIst command is probably what you are looking for.
Mario Sanches
----------
From: Jacob Keller
Let me refine my question (sorry for my lack of clarity): is there a
program that will output the distances between the corresponding ca's
of a superposition on a residue-by-residue basis, and not just a
global RMSD value (doubtless these numbers are part of the
superposition algorithm itself)? I want to plot these values as a
function of residue number to show which parts of the structures
deviate more or less from each other.
Jacob
----------
From: Abhinav Kumar
CCP4's superpose does this.
It produces a table like this, but only for residues that are used in the alignment.
.-------------.------------.-------------.
| Query | Dist.(A) | Target |
|-------------+------------+-------------|
| + A:ARG 10 | |H+ A:GLN 223 |
| + A:ARG 11 | |H+ A:ASN 224 |
|H+ A:LYS 12 | <..6.19..> |H+ A:PRO 225 |
|H+ A:LYS 13 | <--4.03--> |H+ A:ASN 226 |
|H+ A:LYS 14 | <--3.73--> |H+ A:GLN 227 |
|H+ A:GLN 15 | <..4.05..> |H- A:LEU 228 |
|H+ A:LYS 16 | <..3.41..> |H- A:ILE 229 |
|H+ A:GLU 17 | <..1.75..> |H. A:SER 230 |
|H- A:ILE 18 | <++1.97++> |H- A:LEU 231 |
Thanks,
Abhinav
----------
From: Robert Immormino
Hi Jacob,
This sounds like what you get from ddmp (
http://www.csb.yale.edu/userguides/datamanip/ddmp/ddmp_descrip.html ).
Good Luck,
-bob
From: Eric Pettersen
On Nov 28, 2011, at 4:03 PM, Jacob Keller wrote:
If you use the MatchMaker tool in UCSF Chimera to make the superposition, it has an option to show the corresponding sequence alignment. The sequence alignment will have an "RMSD" header running across the top, which is a bar graph of the RMSD values. You can the alignment's Headers->Save... menu item to save the numerical values to a file if you want.
If you already have the structures superimposed on your own, you can use Chimera's Match->Align tool to create a superposition-based sequence alignment, and do the same thing with its RMSD header.
Some links:
Chimera home page: http://www.cgl.ucsf.edu/chimera
alignment tool's RMSD header: http://www.cgl.ucsf.edu/chimera/current/docs/ContributedSoftware/multalignviewer/multalignviewer.html#assessment
--Eric
----------
From: Edward A. Berry
The program:
http://sb20.lbl.gov/berry/for/pdbdist2b.for
does this.
If you run it by the wrapper pdbd2b:
echo 'Find distances greater than threshold between corresponding atoms in 2 PDB files'
echo 'Usage: pdbd2b file1 file2 startres# [thresh]'
pdbdist2b <<eof
$1
$2
$3 10000
$3
$4 3.0
eof
then the format is:
pdbd2b file1 file2 startresiduenumber threshold
(set the threshold to -1 to print all distances)
then it is easy to redirect the output table to a textfile which can be
imported to scalc or excel for plotting.
It is kind of stupid: ignores chain letters, and once it finds
startresiduenumber in both files it proceeds without looking at numbers
or chains, so the atoms have to be in the same order in both files.
But if you have an insertion deletion you can just run twice, starting after the midmatch the second time.
N GLU A 13 N GLU A 13 0.0105
CA GLU A 13 CA GLU A 13 0.0092
CB GLU A 13 CB GLU A 13 0.0186
CG GLU A 13 CG GLU A 13 0.0369
CD GLU A 13 CD GLU A 13 0.0490
OE1 GLU A 13 OE1 GLU A 13 0.0623
OE2 GLU A 13 OE2 GLU A 13 0.0494
C GLU A 13 C GLU A 13 0.0231
O GLU A 13 O GLU A 13 0.0648
N THR A 14 N THR A 14 0.0156
CA THR A 14 CA THR A 14 0.0112
CB THR A 14 CB THR A 14 0.0160
OG1 THR A 14 OG1 THR A 14 0.0164
CG2 THR A 14 CG2 THR A 14 0.0273
C THR A 14 C THR A 14 0.0157
-
written for g77 but seems to compile with gfortran
I need to look in Dave's calendar prog to remind me how to get
command lines parameters in fortran - then can dispense with the wrapper.
----------
From: Tim Gruene
Dear Jacob,
Coot prints this information to the terminal, so you can start coot and
'tee' its output into a file.
Tim
- --
- --
----------
From: Jacob Keller
Dear Crystallographers,
Thanks very much for all who responded to my request. Below is a
compiled list of the various ways to skin this crystallographic cat!
Jacob
-Moleman ca plot distance-CCP4's superpose-distance matrix analysis,
e.g. http://www.csb.yale.edu/userguides/datamanip/ddmp/ddmp_descrip.html,
or other programs-Lsqkab, with the "deltas" card-If you use the
you want. OR If you already have the structures superimposed on your
RMSD header.-Coot prints this information to the terminal, so you can
start coot and 'tee' its output into a file.-Various fortran/homebrew
programs available from individuals-LGA server:
http://proteinmodel.org/AS2TS/LGA/lga.html
----------
From: Eleanor Dodson
Dont you get that from SSM and lsqkab (you have to turn on - list atom distances or some such in the GUI)
Eleanor
----------
From: Gerard DVD Kleywegt
You mean to produce something like this? http://xray.bmc.uu.se/usf/pics/distplot_1chr.gif
That can be done with LSQMAN - http://xray.bmc.uu.se/usf/lsqman_man.html#S65
Of course, there are lots of other interesting metrics and plots to assess structural differences between two models - http://xray.bmc.uu.se/usf/lsqman_man.html#H17
One of my favourite ones is the CD plot, for comparing multiple models, which looks a bit like a gel: http://xray.bmc.uu.se/usf/pics/cdplot_1ldn.gif
For other options, see http://xray.bmc.uu.se/usf/lsqman_man.html#H19
--DVD
******************************************************************
Gerard J. Kleywegt
No comments:
Post a Comment