Discussion:
[gmx-users] g_traj
Alberto.Imparato
2003-07-30 00:21:01 UTC
Permalink
Hi all
I'm simulating a system with 1024 DPPC
and 23552 SOL.
I want to plot the coordinates of my system, so I run
g_traj -f MD -s MD -ox
after selecting the group I want to plot (1=DPPC),
the machine gets very slow and then the program
exits with a "killed" message, without writing
anything in the coord.xvg file.

In my mdp file I set
nstxout = 500


Has anybody an idea about this problem?
Bye
A.
Andrey V Golovin
2003-07-30 00:26:01 UTC
Permalink
Hello Alberto,

Tuesday, July 29, 2003, 9:20:23 PM, you wrote:
Hi I think you are asking too much ...

AInii> g_traj -f MD -s MD -ox
AInii> after selecting the group I want to plot (1=DPPC),

That mean that g_traj must write coordinates (3 column) for ALL atoms
in DPPC in all points of time ... use -com or ....

AInii> the machine gets very slow and then the program
AInii> exits with a "killed" message, without writing
AInii> anything in the coord.xvg file.

AInii> In my mdp file I set
AInii> nstxout = 500


AInii> Has anybody an idea about this problem?
AInii> Bye
AInii> A.



AInii> _______________________________________________
AInii> gmx-users mailing list
AInii> gmx-users at gromacs.org
AInii> http://www.gromacs.org/mailman/listinfo/gmx-users
AInii> Please don't post (un)subscribe requests to the list. Use the
AInii> www interface or send it to gmx-users-request at gromacs.org.
--
Best regards,
Andrey mailto:golovin at genebee.msu.su
Osmany Guirola Cruz
2003-07-30 01:02:01 UTC
Permalink
in the site http://moose.bio.ucalgary.ca i find PDB bilayers and ITP files
whit editconf i have the gro file but not the top file
when i use pdb2gmx , i obtain MANY warnings
i need to know how to convert the itp file to a top file
What should i do?
David
2003-07-30 01:06:01 UTC
Permalink
Post by Osmany Guirola Cruz
in the site http://moose.bio.ucalgary.ca i find PDB bilayers and ITP files
whit editconf i have the gro file but not the top file
when i use pdb2gmx , i obtain MANY warnings
i need to know how to convert the itp file to a top file
What should i do?
I'm sorry but you have to read the manual, chapter 5.
Post by Osmany Guirola Cruz
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
--
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, Dept. of Cell and Molecular Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Itamar Kass
2003-07-30 02:15:01 UTC
Permalink
Post by Osmany Guirola Cruz
in the site http://moose.bio.ucalgary.ca i find PDB bilayers and ITP files
whit editconf i have the gro file but not the top file
when i use pdb2gmx , i obtain MANY warnings
i need to know how to convert the itp file to a top file
What should i do?
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Hi Osmany, I can give you the answer I got from Bert de Groot when I asked
the same qustion:


you don't need pdb2gmx for the lipids. Just include the ffgmx.itp from the
ffgmx_lipids.tar.gz (which also defines the params from lipid.itp, so that
you don't need to include lipid.itp anymore). Then include dmpc.itp and you're
ready to run.

your topology could look something like this:

#include "./ffgmx.itp"
#include "dmpc.itp"

; Include water topology
#ifdef FLEX_SPC
#include "flexspc.itp"
#else
#include "spc.itp"
#endif

[ system ]
; Name
DMPC in water

[ molecules ]
; Compound #mols
DMPC 271
SOL 4238



********************************************
Computers are like airconditioners... They don't work well with Windows
open.
********************************************

===========================================
| Itamar Kass
| The Alexander Silberman
| Institute of Life Sciences
| Department of Biological Chemistry
| The Hebrew University, Givat-Ram
| Jerusalem, 91904, Israel
| Tel: +972-(0)2-6585146
| Fax: +972-(0)2-6584329
| Email: ikass at cc.huji.ac.il
| Group Homepage: http://www.ls.huji.ac.il/~membranelab/
============================================

-------------------------------------------------
This mail sent through IMP: http://horde.org/imp/
Osmany Guirola Cruz
2003-07-30 19:46:01 UTC
Permalink
What time is necesary for cooling a sistem? everybody say SLOWLY .
Osmany Guirola Cruz
2003-07-30 20:04:01 UTC
Permalink
how i do a simulation(simulated annealing)?. i need modify the weight of
the bond angles and bond
length fot the simulation how?
Post by Osmany Guirola Cruz
What time is necesary for cooling a sistem? everybody say SLOWLY .
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Xavier Periole
2003-07-30 23:53:01 UTC
Permalink
Post by Osmany Guirola Cruz
What time is necesary for cooling a sistem? everybody say SLOWLY .
Depends the type of system, its size, the purpose of the simulated
annealing.
But yes the slowest the better. At the origin the simulated annealing was
coupled to Monte Carlo simulation and in THEORY if you decrease the
temperature by steps infinitey small you get to the global minimum. Check
in a book or some papers using simulated annealing.

XAvier
Osmany Guirola Cruz
2003-07-31 01:09:01 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/733037f4/attachment.html>
Xavier Periole
2003-07-31 01:23:00 UTC
Permalink
I guess you have been reading papers from NMR procedures to
generate their models corresponding to the data. If that what you
are doing It would be easier to get a program that NMR people
use, is is probably implemented
It seeems difficult to do that automatically in gromacs, at least for me.
I guess you that could modify the strength of the constant on the angles
and bonds in the parameter file and do that for each temperature you
need but that's gonna to be a pain in the ...
I don't know what you are doing but I can assure you that at 10000 K
the system is pretty flexible. Try to simulate it for 10 ps at 10000 K
and look at the trajectory !! It should convice you. But I don't know if
it is relevant for your problem.

XAvier


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/12fbcb31/attachment.html>
Osmany Guirola Cruz
2003-07-31 01:44:00 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/0b3afe7f/attachment.html>
David
2003-07-31 02:21:01 UTC
Permalink
Thanks again Xavier,
I have a peptide whit unknown structure, whit "modeler" i get 10
diferents structures of my peptide, and then.....
i use SA whith the ten models to obtain a "superminimized structure"
but my final structures are NOT similar
:-( . i have the idea that my simulations converge to a unique
structure (my SA = "1000K" to 300K) you say 10 000 i will probe
with 10000 to see what happens. ???????
I wouldn't use SA in this case, but rather simulated the peptides in
solvent (each of the ten with exactly the same number of solvent
molecules and box size) and then first let it relax and then do a short
free simulation to measure the energy (Epot including solvent). Your
lowest energy is the most likely structure, although the forcefield can
be wrong.
Post by Xavier Periole
I guess you have been reading papers from NMR procedures to
generate their models corresponding to the data. If that what you
are doing It would be easier to get a program that NMR people
use, is is probably implemented
It seeems difficult to do that automatically in gromacs, at least for me.
I guess you that could modify the strength of the constant on the angles
and bonds in the parameter file and do that for each temperature you
need but that's gonna to be a pain in the ...
I don't know what you are doing but I can assure you that at 10000 K
the system is pretty flexible. Try to simulate it for 10 ps at 10000 K
and look at the trajectory !! It should convice you. But I don't know if
it is relevant for your problem.
XAvier
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post
(un)subscribe requests to the list. Use the www interface or send it
to gmx-users-request at gromacs.org.
--
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, Dept. of Cell and Molecular Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Osmany Guirola Cruz
2003-07-31 02:30:02 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/ff1fe0c6/attachment.html>
Xavier Periole
2003-07-31 02:36:04 UTC
Permalink
1- editconf -f p.gro -o p.gro -bt cubic -c -d 0.7
2- genbox -cp p.gro -cs spc216.gro -o p.gro -p p.top
my peptides have diferent structure , and then my BOX
and my number of H2O are diferent, my question is
hOW COULD I GENERATE exactly the same number of solvent
molecules and box size.

include the solvent in your SA. I did that too with gromacs. It works
fine. And don't need a big box for 11 residues.

The implicit solvent SA I used before was with CHARMM implemented
in the lab by. S. Hassan. Can write him a message at
mago at inka.mssm.edu you'll need CHARMM though

XAvier
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/37660599/attachment.html>
Osmany Guirola Cruz
2003-07-31 03:03:01 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/7cf6a067/attachment.html>
Xavier Periole
2003-07-31 03:20:01 UTC
Permalink
1) You can force "genbox" to insert extramolecules of solvent in your box.
And equilibrate the whole thing, but your peptide conformations have
been generated without solvent, right ! HOw good is that ? I am not sure.

2) You can re-do your SAs with explicit solvent ! You beggin with a conformation
of your peptide. You heat it up to 1000 K run for 50 ps and begin to cool down.
Much longer, but the size of the system is small. You can compare the
structures at the end of the 50ps at 1000K and see if they are different enough.
If they are you are can assume that you are going to explore enough space.
ideally. But I'd do more SAs.

Good luck
XAvier



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/899ed636/attachment.html>
Osmany Guirola Cruz
2003-07-31 20:30:01 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030731/b014b8fe/attachment.html>
Xavier Periole
2003-08-01 02:04:00 UTC
Permalink
1) How can i use genbox to insert extramolecules of solvent?



a simple genbox -h gives you the answer. But basically something like :

genbox -cp protein.gro -cs spc216.gro -nmol <number of molecules you want>
-try ......

take a look.

XAvier


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030731/a7358700/attachment.html>
Osmany Guirola Cruz
2003-08-01 02:31:01 UTC
Permalink
Thanks Xaviier
Post by Xavier Periole
1) How can i use genbox to insert extramolecules of solvent?
genbox -cp protein.gro -cs spc216.gro -nmol <number of molecules you want>
-try ......
take a look.
XAvier
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030731/435f7fd2/attachment.html>
David
2003-07-31 03:06:02 UTC
Permalink
Post by Xavier Periole
1- editconf -f p.gro -o p.gro -bt cubic -c -d 0.7
2- genbox -cp p.gro -cs spc216.gro -o p.gro -p p.top
my peptides have diferent structure , and then my BOX
and my number of H2O are diferent, my question is.
hOW COULD I GENERATE exactly the same number of solvent
molecules and box size.
thanks
take the smallest number of any you get from genbox, and take away
molecules manually from the other ones.
Post by Xavier Periole
Post by David
Thanks again Xavier,
I have a peptide whit unknown structure, whit "modeler" i get 10
diferents structures of my peptide, and then.....
i use SA whith the ten models to obtain a "superminimized structure"
but my final structures are NOT similar
:-( . i have the idea that my simulations converge to a unique
structure (my SA = "1000K" to 300K) you say 10 000 i will probe
with 10000 to see what happens. ???????
I wouldn't use SA in this case, but rather simulated the peptides in
solvent (each of the ten with exactly the same number of solvent
molecules and box size) and then first let it relax and then do a short
free simulation to measure the energy (Epot including solvent). Your
lowest energy is the most likely structure, although the forcefield can
be wrong.
Post by Xavier Periole
I guess you have been reading papers from NMR procedures to
generate their models corresponding to the data. If that what you
are doing It would be easier to get a program that NMR people
use, is is probably implemented
It seeems difficult to do that automatically in gromacs, at least for me.
I guess you that could modify the strength of the constant on the angles
and bonds in the parameter file and do that for each temperature you
need but that's gonna to be a pain in the ...
I don't know what you are doing but I can assure you that at 10000 K
the system is pretty flexible. Try to simulate it for 10 ps at 10000 K
and look at the trajectory !! It should convice you. But I don't know if
it is relevant for your problem.
XAvier
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post
(un)subscribe requests to the list. Use the www interface or send it
to gmx-users-request at gromacs.org.
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post
(un)subscribe requests to the list. Use the www interface or send it
to gmx-users-request at gromacs.org.
--
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, Dept. of Cell and Molecular Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Osmany Guirola Cruz
2003-07-31 02:08:01 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/b1807efd/attachment.html>
Xavier Periole
2003-07-31 02:30:04 UTC
Permalink
Yop, sorry I meant 1000 K. 10000 K is way too much.

Marco is right. With 11 residues it is unlikely that your peptide has a unique
conformation in solvent. It probably explore different conformations with diffenrent
probabilities.

How did you generate your 10 structures with Modeler. I thought it needed a
template !! And what isthe point of doing an SA on a specific conformation ?
You loose it at 1000K anyway !!
I did some thing similar where I generated 100 conf from SA with an implicit
solvent (faster) and after that I made clusters of them.

XAvier

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/6f07f1aa/attachment.html>
Osmany Guirola Cruz
2003-07-31 03:01:01 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/d2a02690/attachment.html>
Xavier Periole
2003-07-30 00:32:01 UTC
Permalink
traj file too big for the computer to load it !
David
2003-07-30 00:59:00 UTC
Permalink
Post by Alberto.Imparato
Hi all
I'm simulating a system with 1024 DPPC
and 23552 SOL.
I want to plot the coordinates of my system, so I run
g_traj -f MD -s MD -ox
after selecting the group I want to plot (1=DPPC),
the machine gets very slow and then the program
exits with a "killed" message, without writing
anything in the coord.xvg file.
In my mdp file I set
nstxout = 500
Has anybody an idea about this problem?
How big is the trj/xtc file?

Otherwise the program is not very complicated, so have a look at the
source code or try to run it in the debugger.
The message you mention usually means out of memory.
Post by Alberto.Imparato
Bye
A.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
--
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, Dept. of Cell and Molecular Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Marc Ceruso
2003-07-31 02:10:01 UTC
Permalink
Thanks again Xavier,
I have a peptide whit unknown structure, whit "modeler" i get 10
diferents structures of my peptide, and then.....
i use SA whith the ten models to obtain a "superminimized structure" but
my final structures are NOT similar
How long is the peptide? Why do you think that your structures should be
similar? In any case, if the peptide is small let's say 1-40 amino acids
chances are that its structure in solution is not unique, but in any case
the similiraty within your ensemble will depend on the strength of the
restraints used to generate the structures. If the homology of your
peptide/protein is low the constraints will be loose reflecting the lack
of empirical knowlege regarding your sequence, in others reflecting the
uncertainty in the structure.
:-( . i have the idea that my simulations converge to a unique structure
(my SA = "1000K" to 300K) you say 10 000 i will probe
with 10000 to see what happens. ???????
I guess you have been reading papers from NMR procedures to
generate their models corresponding to the data. If that what you
are doing It would be easier to get a program that NMR people
use, is is probably implemented
It seeems difficult to do that automatically in gromacs, at least
for me.
I guess you that could modify the strength of the constant on the
angles
and bonds in the parameter file and do that for each temperature
you
need but that's gonna to be a pain in the ...
I don't know what you are doing but I can assure you that at 10000
K
the system is pretty flexible. Try to simulate it for 10 ps at
10000 K
and look at the trajectory !! It should convice you. But I don't
know if
it is relevant for your problem.
XAvier
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Osmany Guirola Cruz
2003-07-31 02:16:03 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/4815d6d3/attachment.html>
Kay Gottschalk
2003-07-31 02:15:01 UTC
Permalink
At 10000K you can only do torsion angle dynamics, I think. Your box is
just exploding. I am not sure that you will find a sa scheme with which
your peptide models converge. Do you have any experimental restraints?
How big is the peptide?
At ?10000K my simulation does not work ?i have errors like this.
Back Off! I just backed up ener.edr to ./#ener.edr.3#
starting mdrun 'Protein in water'
5000 steps,???? 10.0 ps.
Back Off! I just backed up peptido_md.trr to ./#peptido_md.trr.3#
step 90, remaining runtime:?? 269 s
Back Off! I just backed up step100.pdb to ./#step100.pdb.1#
Wrote pdb files with previous and current coordinates
step 100, remaining runtime:?? 291 s
Back Off! I just backed up step103.pdb to ./#step103.pdb.1#
Back Off! I just backed up step104.pdb to ./#step104.pdb.2#
Wrote pdb files with previous and current coordinates
Warning: Only triclinic boxes with the first vector parallel to the
x-axis and the second vector in the xy-plane are supported.
??????????? Box[??? 0]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 1]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 2]={???????? nan,????????? nan,????????? nan}
???????? Can not fix pbc.
Warning: Only triclinic boxes with the first vector parallel to the
x-axis and the second vector in the xy-plane are supported.
??????????? Box[??? 0]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 1]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 2]={???????? nan,????????? nan,????????? nan}
???????? Can not fix pbc.
Warning: Only triclinic boxes with the first vector parallel to the
x-axis and the second vector in the xy-plane are supported.
??????????? Box[??? 0]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 1]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 2]={???????? nan,????????? nan,????????? nan}
???????? Can not fix pbc.
Warning: Only triclinic boxes with the first vector parallel to the
x-axis and the second vector in the xy-plane are supported.
??????????? Box[??? 0]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 1]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 2]={???????? nan,????????? nan,????????? nan}
???????? Can not fix pbc.
Warning: Only triclinic boxes with the first vector parallel to the
x-axis and the second vector in the xy-plane are supported.
??????????? Box[??? 0]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 1]={???????? nan,????????? nan,????????? nan}
??????????? Box[??? 2]={???????? nan,????????? nan,????????? nan}
???????? Can not fix pbc.
Fatal error: ci = -1 should be in 0 .. -1 [FILE nsgrid.c, LINE 210]
Ok, thanks Xavier, but i read in some papers that they modify? bond
angles and bond
length, ?for give more "movility" to the system and guarantee a great
SA. How could ido this with
gromacs
What time is necesary for cooling a sistem? everybody say SLOWLY .
Depends the type of system, its size, the purpose of the simulated
annealing.
But yes the slowest the better. At the origin the simulated annealing
was
coupled to Monte Carlo simulation and in THEORY if you decrease the
temperature by steps infinitey small you get to the global minimum.
Check
in a book or some papers using simulated annealing.
XAvier
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post
(un)subscribe requests to the list. Use the www interface or send it
to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 4762 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/24fe5651/attachment.bin>
Osmany Guirola Cruz
2003-07-31 02:19:00 UTC
Permalink
my peptide have 11 amino acids and it is cyclic
Post by Kay Gottschalk
At 10000K you can only do torsion angle dynamics, I think. Your box is
just exploding. I am not sure that you will find a sa scheme with
which your peptide models converge. Do you have any experimental
restraints? How big is the peptide?
At 10000K my simulation does not work i have errors like this.
Back Off! I just backed up ener.edr to ./#ener.edr.3#
starting mdrun 'Protein in water'
5000 steps, 10.0 ps.
Back Off! I just backed up peptido_md.trr to ./#peptido_md.trr.3#
step 90, remaining runtime: 269 s
Back Off! I just backed up step100.pdb to ./#step100.pdb.1#
Wrote pdb files with previous and current coordinates
step 100, remaining runtime: 291 s
Back Off! I just backed up step103.pdb to ./#step103.pdb.1#
Back Off! I just backed up step104.pdb to ./#step104.pdb.2#
Wrote pdb files with previous and current coordinates
Warning: Only triclinic boxes with the first vector parallel to
the x-axis and the second vector in the xy-plane are supported.
Box[ 0]={ nan, nan, nan}
Box[ 1]={ nan, nan, nan}
Box[ 2]={ nan, nan, nan}
Can not fix pbc.
Warning: Only triclinic boxes with the first vector parallel to
the x-axis and the second vector in the xy-plane are supported.
Box[ 0]={ nan, nan, nan}
Box[ 1]={ nan, nan, nan}
Box[ 2]={ nan, nan, nan}
Can not fix pbc.
Warning: Only triclinic boxes with the first vector parallel to
the x-axis and the second vector in the xy-plane are supported.
Box[ 0]={ nan, nan, nan}
Box[ 1]={ nan, nan, nan}
Box[ 2]={ nan, nan, nan}
Can not fix pbc.
Warning: Only triclinic boxes with the first vector parallel to
the x-axis and the second vector in the xy-plane are supported.
Box[ 0]={ nan, nan, nan}
Box[ 1]={ nan, nan, nan}
Box[ 2]={ nan, nan, nan}
Can not fix pbc.
Warning: Only triclinic boxes with the first vector parallel to
the x-axis and the second vector in the xy-plane are supported.
Box[ 0]={ nan, nan, nan}
Box[ 1]={ nan, nan, nan}
Box[ 2]={ nan, nan, nan}
Can not fix pbc.
Fatal error: ci = -1 should be in 0 .. -1 [FILE nsgrid.c, LINE 210]
Ok, thanks Xavier, but i read in some papers that they modify
bond angles and bond
length, for give more "movility" to the system and guarantee a
great SA. How could ido this with
gromacs
What time is necesary for cooling a sistem? everybody say SLOWLY .
Depends the type of system, its size, the purpose of the simulated
annealing.
But yes the slowest the better. At the origin the simulated
annealing was
coupled to Monte Carlo simulation and in THEORY if you decrease the
temperature by steps infinitey small you get to the global
minimum. Check
in a book or some papers using simulated annealing.
XAvier
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________ gmx-users mailing
list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users Please don't
post (un)subscribe requests to the list. Use the www interface or
send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/b980362b/attachment.html>
Marc Ceruso
2003-07-31 02:25:01 UTC
Permalink
my peptide have 11 aminoacids and it is cyclic.
So unless you have NMR data, what you're trying to do is predict the
structure of your peptdide from scratch?
In the latter case at best you will find ensembles of distinct conformers,
Maybe the papers van Gunsteren's group on the molecular dynamics of a
small peptide will help you :

Proteins. 2001 Apr 1;43(1):45-56.
Entropy calculations on a reversibly folding peptide: changes in solute
free energy cannot explain folding behavior.
Schafer H, Daura X, Mark AE, van Gunsteren WF


Marc
Kay Gottschalk
2003-07-31 02:26:02 UTC
Permalink
It might very well be that your system is way to flexible; you'd need
experimental (=NMR) restraints, I'd guess. Even a cyclic hexapeptide is
very flexible and basically improbable to predict correctly.
my peptide have 11 aminoacids and it is cyclic.
Thanks again Xavier,
I have a peptide whit unknown structure, whit "modeler" i get 10
diferents structures of my peptide, and then.....
i use SA whith the ten models to obtain a "superminimized structure"
but
my final structures are NOT similar
How long is the peptide? Why do you think that your structures should
be
similar? In any case, if the peptide is small let's say 1-40 amino
acids
chances are that its structure in solution is not unique, but in any
case
the similiraty within your ensemble will depend on the strength of the
restraints used to generate the structures. If the homology of your
peptide/protein is low the constraints will be loose reflecting the
lack
of empirical knowlege regarding your sequence, in others reflecting the
uncertainty in the structure.
:-( . i have the idea that my simulations converge to a unique
structure
(my SA = "1000K" to 300K) you say 10 000 i will probe
with 10000 to see what happens. ???????
I guess you have been reading papers from NMR procedures to
generate their models corresponding to the data. If that what you
are doing It would be easier to get a program that NMR people
use, is is probably implemented
It seeems difficult to do that automatically in gromacs, at least
for me.
I guess you that could modify the strength of the constant on the
angles
and bonds in the parameter file and do that for each temperature
you
need but that's gonna to be a pain in the ...
I don't know what you are doing but I can assure you that at 10000
K
the system is pretty flexible. Try to simulate it for 10 ps at
10000 K
and look at the trajectory !! It should convice you. But I don't
know if
it is relevant for your problem.
XAvier
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 3473 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/b1d5b266/attachment.bin>
Osmany Guirola Cruz
2003-07-31 02:37:01 UTC
Permalink
I don't have NMR information i am trying to predict the structure of my
peptdide from scratch
, like Marc Ceruso says in another mail.
Post by Kay Gottschalk
It might very well be that your system is way to flexible; you'd need
experimental (=NMR) restraints, I'd guess. Even a cyclic hexapeptide
is very flexible and basically improbable to predict correctly.
my peptide have 11 aminoacids and it is cyclic.
Thanks again Xavier,
I have a peptide whit unknown structure, whit "modeler" i get 10
diferents structures of my peptide, and then.....
i use SA whith the ten models to obtain a "superminimized
structure" but
my final structures are NOT similar
How long is the peptide? Why do you think that your structures
should be
similar? In any case, if the peptide is small let's say 1-40 amino
acids
chances are that its structure in solution is not unique, but in
any case
the similiraty within your ensemble will depend on the strength of the
restraints used to generate the structures. If the homology of your
peptide/protein is low the constraints will be loose reflecting
the lack
of empirical knowlege regarding your sequence, in others
reflecting the
uncertainty in the structure.
:-( . i have the idea that my simulations converge to a unique
structure
(my SA = "1000K" to 300K) you say 10 000 i will probe
with 10000 to see what happens. ???????
I guess you have been reading papers from NMR procedures to
generate their models corresponding to the data. If that what you
are doing It would be easier to get a program that NMR people
use, is is probably implemented
It seeems difficult to do that automatically in gromacs, at least
for me.
I guess you that could modify the strength of the constant on the
angles
and bonds in the parameter file and do that for each temperature
you
need but that's gonna to be a pain in the ...
I don't know what you are doing but I can assure you that at 10000
K
the system is pretty flexible. Try to simulate it for 10 ps at
10000 K
and look at the trajectory !! It should convice you. But I don't
know if
it is relevant for your problem.
XAvier
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/d0faeb3d/attachment.html>
Kay Gottschalk
2003-07-31 02:27:00 UTC
Permalink
If you really want SA, try the program CNS - it has a nice SA
implementation, although you'd have to do the calculations in vacuo
first, I'd guess. After the SA in Vacuo (as a minimization...) you
could switch to Gromacs and do a free simulation in water.
Post by David
Thanks again Xavier,
I have a peptide whit unknown structure, whit "modeler" i get 10
diferents structures of my peptide, and then.....
i use SA whith the ten models to obtain a "superminimized structure"
but my final structures are NOT similar
:-( . i have the idea that my simulations converge to a unique
structure (my SA = "1000K" to 300K) you say 10 000 i will probe
with 10000 to see what happens. ???????
I wouldn't use SA in this case, but rather simulated the peptides in
solvent (each of the ten with exactly the same number of solvent
molecules and box size) and then first let it relax and then do a short
free simulation to measure the energy (Epot including solvent). Your
lowest energy is the most likely structure, although the forcefield can
be wrong.
Post by Xavier Periole
I guess you have been reading papers from NMR procedures to
generate their models corresponding to the data. If that what you
are doing It would be easier to get a program that NMR people
use, is is probably implemented
It seeems difficult to do that automatically in gromacs, at least
for me.
I guess you that could modify the strength of the constant on the
angles
and bonds in the parameter file and do that for each temperature you
need but that's gonna to be a pain in the ...
I don't know what you are doing but I can assure you that at 10000 K
the system is pretty flexible. Try to simulate it for 10 ps at 10000
K
and look at the trajectory !! It should convice you. But I don't
know if
it is relevant for your problem.
XAvier
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users Please don't post
(un)subscribe requests to the list. Use the www interface or send it
to gmx-users-request at gromacs.org.
--
Groeten, David.
_______________________________________________________________________
_
Dr. David van der Spoel, Dept. of Cell and Molecular Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
Kay Gottschalk
2003-07-31 02:36:02 UTC
Permalink
There is a lot of literature about calculation of cyclic peptides by
the group of Kessler in Munich. They do similar things to what Xavier
describes, and use the clusters as input for free MD in water. But they
normally do have NMR restraints...
Post by Xavier Periole
?
?
Yop, sorry I meant 1000 K. 10000 K is way too much.
?
Marco is right. With 11 residues it is unlikely that your peptide has
a unique
conformation in solvent. It probably explore different conformations
with diffenrent
probabilities.
?
How did you generate your 10 structures with Modeler. I thought it
needed a
template !! And what isthe point of doing an SA on a specific
conformation ?
You loose it at 1000K anyway !!
I did some thing similar where I generated 100 conf from SA with an
implicit
solvent (faster) and after that I made?clusters of them.
?
XAvier
?
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 1707 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030730/451b984d/attachment.bin>
Kay Gottschalk
2003-07-31 13:22:01 UTC
Permalink
I'd do much more than ten structures. Do as many as you can in a
reasonable time and afterwards cluster the results! You cannot expect
your system to converge so easily...
that was the problem in my simulation at 1000K my ten final structures
are quite diferent after the SA.
?
?
Yop, sorry I meant 1000 K. 10000 K is way too much.
?
Marco is right. With 11 residues it is unlikely that your peptide has
a unique
conformation in solvent. It probably explore different conformations
with diffenrent
probabilities.
?
How did you generate your 10 structures with Modeler. I thought it
needed a
template !! And what isthe point of doing an SA on a specific
conformation ?
You loose it at 1000K anyway !!
I did some thing similar where I generated 100 conf from SA with an
implicit
solvent (faster) and after that I made?clusters of them.
?
XAvier
?
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 1772 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030731/3d84d7e1/attachment.bin>
Osmany Guirola Cruz
2003-07-31 19:17:01 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030731/95b41957/attachment.html>
Marc Ceruso
2003-07-31 20:18:01 UTC
Permalink
Hi-
Your results are not surprizing. But it seems to me that you would be
happier if you had one single structure for you peptide. Even if it were
that the peptide does have a unique stable conformation the search for it
would be to find a protocol/forcefield that would accurately determine the
correct fold of your peptide: protein folding problem. However, small
peptides do not have a unique
structure in solution. Numerous NMR studies on many peptides have shown
that the peptides (in general) are in equilibirium between different
forms. Take a look at the enkephalin (Scheraga) studies, or some of Ernst
(NMR) studies.
If you have experimental data on your peptide, You could run classical
MD simulations (see Daura and van Gunsteren) for a very long time
(tens/hundreds of ns)and evaluate
your computational results against the NMR data. That would be the measure
of convergence.
I hope this helps
Marco



On Thu, 31 Jul 2003, Osmany
i have 10 simulations , i do g_cluster to each one and then i do g_confrm
whith the most populated clusters
1 2 3 4
5 6 7 8
9 10
1 *
2 0.255333 *
3 0.384980 0.268161 *
4 0.363748 0.356637 0.312702 *
5 0.364454 0.291378 0.168395 0.360910 *
6 0.304329 0.362637 0.352839 0.224768 0.350295 *
7 0.395214 0.427625 0.373925 0.340693 0.329041 0.264797 *
8 0.395720 0.347483 0.315446 0.462918 0.295911 0.384215 0.404703
9 ? ? ? ? ? ? ?
10 0.269739 0.224798 0.375789 0.433382 0.349531 0.403248 0.452217
0.382978
I'd do much more than ten structures. Do as many as you can
in a reasonable time and afterwards cluster the results! You
cannot expect your system to converge so easily...
On Thursday, July 31, 2003, at 02:00 AM, Osmany Guirola Cruz
that was the problem in my simulation at 1000K my
ten final structures are quite diferent after the
SA.
Yop, sorry I meant 1000 K. 10000 K is way too
much.
Marco is right. With 11 residues it is unlikely
that your peptide has a unique
conformation in solvent. It probably explore
different conformations with diffenrent
probabilities.
How did you generate your 10 structures with
Modeler. I thought it needed a
template !! And what isthe point of doing an SA
on a specific conformation ?
You loose it at 1000K anyway !!
I did some thing similar where I generated 100
conf from SA with an implicit
solvent (faster) and after that I made clusters
of them.
XAvier
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Osmany Guirola Cruz
2003-07-31 20:53:00 UTC
Permalink
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030731/4ede5350/attachment.html>
Kay Gottschalk
2003-07-31 21:00:01 UTC
Permalink
To be honest, I think that will not work. You will probably have to
deal with ensembles of probable (=low-energy) structures. To get decent
ensembles, you need much more than 10 structures. If you heat up the
structures to 1000K, the start conformation shouldn't be that important
any more. I'd calculate on the order of 100 or more conformations and
see whether or not I can detect any clusters.
Thanks Marc
But if i have 10 diferent structures for the same peptide and ?i heat
it to 1000K and then cool it to 300K
to each one , i supose that my 10 final structures must be similar. My
idea was to find a unique structure for my peptides
an then do MD.
Hi-
Your results are not surprizing. But it seems to me that you would be
happier if you had one single structure for you peptide. Even if it
were
that the peptide does have a unique stable conformation the search for
it
would be to find a protocol/forcefield that would accurately determine
the
correct fold of your peptide: protein folding problem. However, small
peptides do not have a unique
structure in solution. Numerous NMR studies on many peptides have shown
that the peptides (in general) are in equilibirium between different
forms. Take a look at the enkephalin (Scheraga) studies, or some of
Ernst
(NMR) studies.
If you have experimental data on your peptide, You could run classical
MD simulations (see Daura and van Gunsteren) for a very long time
(tens/hundreds of ns)and evaluate
your computational results against the NMR data. That would be the
measure
of convergence.
I hope this helps
Marco
On Thu, 31 Jul 2003, Osmany
i have 10 simulations , i do g_cluster to each one and then i do
g_confrm
whith the most populated clusters
1 2 3 4
5 6 7 8
9 10
1 *
2 0.255333 *
3 0.384980 0.268161 *
4 0.363748 0.356637 0.312702 *
5 0.364454 0.291378 0.168395 0.360910 *
6 0.304329 0.362637 0.352839 0.224768 0.350295 *
7 0.395214 0.427625 0.373925 0.340693 0.329041 0.264797 *
8 0.395720 0.347483 0.315446 0.462918 0.295911 0.384215
0.404703
9 ? ? ? ? ? ? ?
10 0.269739 0.224798 0.375789 0.433382 0.349531 0.403248
0.452217
0.382978
I'd do much more than ten structures. Do as many as you can
in a reasonable time and afterwards cluster the results! You
cannot expect your system to converge so easily...
On Thursday, July 31, 2003, at 02:00 AM, Osmany Guirola Cruz
that was the problem in my simulation at 1000K my
ten final structures are quite diferent after the
SA.
Yop, sorry I meant 1000 K. 10000 K is way too
much.
Marco is right. With 11 residues it is unlikely
that your peptide has a unique
conformation in solvent. It probably explore
different conformations with diffenrent
probabilities.
How did you generate your 10 structures with
Modeler. I thought it needed a
template !! And what isthe point of doing an SA
on a specific conformation ?
You loose it at 1000K anyway !!
I did some thing similar where I generated 100
conf from SA with an implicit
solvent (faster) and after that I made clusters
of them.
XAvier
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
_______________________________________________ gmx-users mailing list
gmx-users at gromacs.org http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 5237 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030731/6c08daca/attachment.bin>
Albert Sun
2003-08-01 04:29:01 UTC
Permalink
How can we use other molecules to replace protein and solvent when using genbox?
could we use other .gro files to replace protein.gro and spc216.gro ?





Osmany Guirola Cruz <osmany.guirola at cigb.edu.cu> wrote:
Thanks Xaviier


Xavier Periole wrote:



1) How can i use genbox to insert extramolecules of solvent?



a simple genbox -h gives you the answer. But basically something like :

genbox -cp protein.gro -cs spc216.gro -nmol <number of molecules you want>
-try ......

take a look.

XAvier






---------------------------------
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030731/ed191b80/attachment.html>
Xavier Periole
2003-08-01 04:46:01 UTC
Permalink
Sorry I though it was clear.
Those names are the names of your protein file (or whatever system)
in an empty box or I guess a box filled with water. And the spc216.gro
is the water box used in gromacs to fill up an empty box. You can
probably take your own.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030731/1c31f2ad/attachment.html>
David van der Spoel
2003-08-01 14:38:00 UTC
Permalink
Post by Albert Sun
How can we use other molecules to replace protein and solvent when using genbox?
could we use other .gro files to replace protein.gro and spc216.gro ?
yes, the only restriction is that the solvent should be one residue per
molecule. e.g. dmso.gro could be used. You can also use mixed solvent,
i.e. you can start with an equilibrated water/methanol box and solvate
your protein in that.
Post by Albert Sun
Thanks Xaviier
Post by Xavier Periole
1) How can i use genbox to insert extramolecules of
solvent?
genbox -cp protein.gro -cs spc216.gro -nmol <number of
molecules you want>
-try ......
take a look.
XAvier
______________________________________________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
--
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, Dept. of Cell & Mol. Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Kay Gottschalk
2003-08-04 16:57:01 UTC
Permalink
Hi all,

does anyone know how to use g_rdf angle-dependent? In the manual it's
written that one can define some angle theta (which affects the
averaging), but somehow I can't figure out how to define that angle.
Thanks for your help,
Kay.
David
2003-08-04 17:26:01 UTC
Permalink
Post by Kay Gottschalk
Hi all,
does anyone know how to use g_rdf angle-dependent? In the manual it's
written that one can define some angle theta (which affects the
averaging), but somehow I can't figure out how to define that angle.
Thanks for your help,
Kay.
I think that option has been lost a long while ago...
Post by Kay Gottschalk
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
--
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, Dept. of Cell and Molecular Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Anton Feenstra
2003-08-06 14:47:02 UTC
Permalink
Post by David
I think that option has been lost a long while ago...
Yes - it was found to be completely dysfunctional.
Sorry we forgot to update the manual.... ;-{
--
Groetjes,

Anton
_____________ _______________________________________________________
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610 |
| | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/ |
| | "If You See Me Getting High, Knock Me Down" |
| | (Red Hot Chili Peppers) |
|_____________|_______________________________________________________|
Kay Gottschalk
2003-08-04 19:55:01 UTC
Permalink
Hi all,

does anyone know how to use g_rdf angle-dependent? In the manual it's
written that one can define some angle theta (which affects the
averaging), but somehow I can't figure out how to define that angle.
Thanks for your help,
Kay.
Kay Gottschalk
2003-08-04 20:01:01 UTC
Permalink
Hmm, too bad. Any idea how to do the averaging then? If I want to
calculate the rdf of solvent around a certain surface atom, I will have
to compensate for volume of the protein, I think. I am not really sure
how to do that...
Thanks,
Kay.
Post by David
Post by Kay Gottschalk
Hi all,
does anyone know how to use g_rdf angle-dependent? In the manual it's
written that one can define some angle theta (which affects the
averaging), but somehow I can't figure out how to define that angle.
Thanks for your help,
Kay.
I think that option has been lost a long while ago...
Post by Kay Gottschalk
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
--
Groeten, David.
_______________________________________________________________________
_
Dr. David van der Spoel, Dept. of Cell and Molecular Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
Kay Gottschalk
2003-08-06 14:58:01 UTC
Permalink
No problem with manual:) I'd just like to know how people do the rdf
for single surface atoms. There must be some correction for the protein
volume (which excludes water of course) in the sphere-section for which
the density function is calculated. Is there a clever way to do that in
Gromacs? Actually, I am not sure I am making myself clear here... At
any rate, suggestions would be highly appreciated.
Cheers,
Kay.
Post by Anton Feenstra
Post by David
I think that option has been lost a long while ago...
Yes - it was found to be completely dysfunctional.
Sorry we forgot to update the manual.... ;-{
--
Groetjes,
Anton
_____________ _______________________________________________________
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610 |
| | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/ |
| | "If You See Me Getting High, Knock Me Down" |
| | (Red Hot Chili Peppers) |
|_____________|_______________________________________________________|
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
Anton Feenstra
2003-08-07 14:15:01 UTC
Permalink
No problem with manual:) I'd just like to know how people do the rdf for
single surface atoms. There must be some correction for the protein
volume (which excludes water of course) in the sphere-section for which
the density function is calculated. Is there a clever way to do that in
Gromacs? Actually, I am not sure I am making myself clear here... At any
rate, suggestions would be highly appreciated.
You'll have to do some explaining here ;-(

We've used rdf's (in the past) for comparison with e.g. neutron scattering
for e.g. water. The angle dependence was added to analyze coordination
shell structure in liquid water. So, no surfaces there, and no protein.

What do you mean by 'single surface atom', and what correction for protein
volume should there be? It may be that g_sas has an option to calculate
the protein surface as well as the volume, but it has been upgraded since
I last used it so I'm not sure.
--
Groetjes,

Anton
_____________ _______________________________________________________
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610 |
| | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/ |
| | "If You See Me Getting High, Knock Me Down" |
| | (Red Hot Chili Peppers) |
|_____________|_______________________________________________________|
Kay Gottschalk
2003-08-07 14:34:01 UTC
Permalink
I have read in the literature that people calculate the rdf of water
around single water exposed protein atoms (for example to find the
difference in the hydration shell around carbons and oxygens). I want
to do something comparable to that. Now if I just take g_rdf and one
protein atom which is solvent exposed, I think that g_rdf normalizes
the number of water molecules that are close to my reference atom
according to the number that is expected in a sphere with a certain
radius around the atom with respect to the density of normal water. The
problem is that most of the atoms' environment is not solvent, but
other protein atoms. g_rdf therefore finds much less water molecules
than expected. I believe that I have to compensate for the volume
around my reference atom that is inaccessible to water. And I don't
know how to do it.
Hope my problem is stated more clearly now - if not, complain...
Thanks,
Kay.
Post by Anton Feenstra
Post by Kay Gottschalk
No problem with manual:) I'd just like to know how people do the rdf
for single surface atoms. There must be some correction for the
protein volume (which excludes water of course) in the sphere-section
for which the density function is calculated. Is there a clever way
to do that in Gromacs? Actually, I am not sure I am making myself
clear here... At any rate, suggestions would be highly appreciated.
You'll have to do some explaining here ;-(
We've used rdf's (in the past) for comparison with e.g. neutron
scattering
for e.g. water. The angle dependence was added to analyze coordination
shell structure in liquid water. So, no surfaces there, and no protein.
What do you mean by 'single surface atom', and what correction for
protein
volume should there be? It may be that g_sas has an option to calculate
the protein surface as well as the volume, but it has been upgraded
since
I last used it so I'm not sure.
--
Groetjes,
Anton
_____________ _______________________________________________________
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610 |
| | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/ |
| | "If You See Me Getting High, Knock Me Down" |
| | (Red Hot Chili Peppers) |
|_____________|_______________________________________________________|
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
Christoph Freudenberger
2003-08-07 14:51:01 UTC
Permalink
Hi Kay.

I'm not sure whether I understood you problem correctly, but if you
want to study solvation around parts of a protein in angular dependence
you might want to consinder using spatial distribution functions (SDF).
I have written a tool to calculate SDF's from gmx trajectories.

Tell me if you want to try it. I will provide the source code.
Post by Kay Gottschalk
I have read in the literature that people calculate the rdf of water
around single water exposed protein atoms (for example to find the
difference in the hydration shell around carbons and oxygens). I want
to do something comparable to that. Now if I just take g_rdf and one
protein atom which is solvent exposed, I think that g_rdf normalizes the
number of water molecules that are close to my reference atom according
to the number that is expected in a sphere with a certain radius around
the atom with respect to the density of normal water. The problem is
that most of the atoms' environment is not solvent, but other protein
atoms. g_rdf therefore finds much less water molecules than expected. I
believe that I have to compensate for the volume around my reference
atom that is inaccessible to water. And I don't know how to do it.
Hope my problem is stated more clearly now - if not, complain...
Thanks,
Kay.
Post by Anton Feenstra
Post by Kay Gottschalk
No problem with manual:) I'd just like to know how people do the rdf
for single surface atoms. There must be some correction for the
protein volume (which excludes water of course) in the sphere-section
for which the density function is calculated. Is there a clever way
to do that in Gromacs? Actually, I am not sure I am making myself
clear here... At any rate, suggestions would be highly appreciated.
You'll have to do some explaining here ;-(
We've used rdf's (in the past) for comparison with e.g. neutron
scattering
for e.g. water. The angle dependence was added to analyze coordination
shell structure in liquid water. So, no surfaces there, and no protein.
What do you mean by 'single surface atom', and what correction for
protein
volume should there be? It may be that g_sas has an option to calculate
the protein surface as well as the volume, but it has been upgraded since
I last used it so I'm not sure.
--
Groetjes,
Anton
_____________ _______________________________________________________
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610 |
| | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/ |
| | "If You See Me Getting High, Knock Me Down" |
| | (Red Hot Chili Peppers) |
|_____________|_______________________________________________________|
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the www
interface or send it to gmx-users-request at gromacs.org.
--
Christoph Freudenberger
Abt. Organische Chemie I AK Siehl - Uni Ulm -Tel: ++49-731-502-2785
Kay Gottschalk
2003-08-07 15:31:01 UTC
Permalink
Hi Christoph,

danke fuer das Angebot! I'd like to try it. But perhaps I should
restate my problem to make it (even more:)) clear: The relative
probability of finding a water water molecule at a given distance r is
(approximately as stated in the manual)

g(r) = <N(r)>/(4*pi*r^2*dr*rho), where N(r) is the number of water in a
sperical shell (4*pi*r^2*dr), normalized by the water density rho.
However, this is only true for symmetrical systems. The problem in my
case is that the protein excludes water from certain regions, thus I
have a assymetrical system. Therefore, the spherical shell is not the
correct volume over which to normalize the function.
Cheers,
Kay.
christoph.freudenberger at chemie.uni-ulm.de
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
Anton Feenstra
2003-08-08 14:08:01 UTC
Permalink
Post by Kay Gottschalk
Hi Christoph,
danke fuer das Angebot! I'd like to try it. But perhaps I should restate
my problem to make it (even more:)) clear: The relative probability of
finding a water water molecule at a given distance r is (approximately
as stated in the manual)
g(r) = <N(r)>/(4*pi*r^2*dr*rho), where N(r) is the number of water in a
sperical shell (4*pi*r^2*dr), normalized by the water density rho.
However, this is only true for symmetrical systems. The problem in my
case is that the protein excludes water from certain regions, thus I
have a assymetrical system. Therefore, the spherical shell is not the
correct volume over which to normalize the function.
Hmm, I'm not sure actually how it is implemented. It could be either as a
distribution function, which is normalized to an integral of 1, or it is
normalized to a bulk density (i.e. at 'infinite' or 'large' distances) of 1,
which is taken from your actual distribution, not an 'external' reference
value. In any case, there is no normalization to a 'reference' average water
density (which would also be pressure & temperature dependent, and water
model dependent as well).

In either case, I believe three columns are written in the output (.xvg)
file, first the distance, then the normalized rdf and finally the 'raw'
numbers (if not, it would be straightforward to add to the source code).
If you type 'xmgrace -nxy rdf.xvg', you will see all available plots in
the file.
--
Groetjes,

Anton
_____________ _______________________________________________________
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610 |
| | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/ |
| | "If You See Me Getting High, Knock Me Down" |
| | (Red Hot Chili Peppers) |
|_____________|_______________________________________________________|
David
2003-08-08 14:22:00 UTC
Permalink
Post by Anton Feenstra
Post by Kay Gottschalk
Hi Christoph,
danke fuer das Angebot! I'd like to try it. But perhaps I should restate
my problem to make it (even more:)) clear: The relative probability of
finding a water water molecule at a given distance r is (approximately
as stated in the manual)
g(r) = <N(r)>/(4*pi*r^2*dr*rho), where N(r) is the number of water in a
sperical shell (4*pi*r^2*dr), normalized by the water density rho.
However, this is only true for symmetrical systems. The problem in my
case is that the protein excludes water from certain regions, thus I
have a assymetrical system. Therefore, the spherical shell is not the
correct volume over which to normalize the function.
Hmm, I'm not sure actually how it is implemented. It could be either as a
distribution function, which is normalized to an integral of 1, or it is
normalized to a bulk density (i.e. at 'infinite' or 'large' distances) of 1,
which is taken from your actual distribution, not an 'external' reference
value. In any case, there is no normalization to a 'reference' average water
density (which would also be pressure & temperature dependent, and water
model dependent as well).
In either case, I believe three columns are written in the output (.xvg)
file, first the distance, then the normalized rdf and finally the 'raw'
numbers (if not, it would be straightforward to add to the source code).
If you type 'xmgrace -nxy rdf.xvg', you will see all available plots in
the file.
I don't think there is any reason to renormalize, the normalization is
done based on the number of water molecules (oxygens) in the
computational box, and since your site of interest is partially occluded
by protein the number of bound water molecules will not be the same as
for a particle in solution.

Try to take the integral of the RDF (using Int 4 pi r^2 g(r) from 0 to
the first minimum), that will give you the average number of hydrogen
bonded water molecules.
--
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, Dept. of Cell and Molecular Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Berk Hess
2003-08-08 20:36:02 UTC
Permalink
Post by David
Try to take the integral of the RDF (using Int 4 pi r^2 g(r) from 0 to
the first minimum), that will give you the average number of hydrogen
bonded water molecules.
The -cn option of g_rdf plots exactly this integral from 0 to r as a
function of r,
so you can see the number of bound water molecules as a function of the
distance.

Berk.

_________________________________________________________________
Add photos to your messages with MSN 8. Get 2 months FREE*.
http://join.msn.com/?page=features/featuredemail
Kay Gottschalk
2003-08-08 20:52:01 UTC
Permalink
Thanks, I'll try that!
Cheers,
K.
Post by David
Post by Anton Feenstra
Post by Kay Gottschalk
Hi Christoph,
danke fuer das Angebot! I'd like to try it. But perhaps I should
restate
my problem to make it (even more:)) clear: The relative probability
of
finding a water water molecule at a given distance r is
(approximately
as stated in the manual)
g(r) = <N(r)>/(4*pi*r^2*dr*rho), where N(r) is the number of water
in a
sperical shell (4*pi*r^2*dr), normalized by the water density rho.
However, this is only true for symmetrical systems. The problem in my
case is that the protein excludes water from certain regions, thus I
have a assymetrical system. Therefore, the spherical shell is not the
correct volume over which to normalize the function.
Hmm, I'm not sure actually how it is implemented. It could be either
as a
distribution function, which is normalized to an integral of 1, or it
is
normalized to a bulk density (i.e. at 'infinite' or 'large'
distances) of 1,
which is taken from your actual distribution, not an 'external'
reference
value. In any case, there is no normalization to a 'reference'
average water
density (which would also be pressure & temperature dependent, and
water
model dependent as well).
In either case, I believe three columns are written in the output
(.xvg)
file, first the distance, then the normalized rdf and finally the
'raw'
numbers (if not, it would be straightforward to add to the source
code).
If you type 'xmgrace -nxy rdf.xvg', you will see all available plots
in
the file.
I don't think there is any reason to renormalize, the normalization is
done based on the number of water molecules (oxygens) in the
computational box, and since your site of interest is partially
occluded
by protein the number of bound water molecules will not be the same as
for a particle in solution.
Try to take the integral of the RDF (using Int 4 pi r^2 g(r) from 0 to
the first minimum), that will give you the average number of hydrogen
bonded water molecules.
--
Groeten, David.
_______________________________________________________________________
_
Dr. David van der Spoel, Dept. of Cell and Molecular Biology
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Dr. Kay-E. Gottschalk
Department of Biological Chemistry
Weizmann Institute of Science
Tel: ++972-8-9343639
Herzl St. 1
Rehovot 76100
Israel
Kay Gottschalk
2003-10-02 19:57:01 UTC
Permalink
Hi there,

is there any way not to run g_rdf interactively? what I mean is the
following: I have many (>100) index files, and I want to run g_rdf with
the first group in the index file as reference as the other for the
distribution function. Doing it >100 times by hand is kind of annoying.
I guess there is some line command which can do it, but somehow I
couldn't figure it out.
Thanks for your help,
Kay.
Marcos Villarreal
2003-10-02 20:43:01 UTC
Permalink
you can do it with a bash script
eg:

#!/bin/bash
n=5 #number of rdf to calculate
g_rdf -n -s <<+
$n
$first #the reference group
1
2
3
4
5
+

I hope this helps.
Saludos,
Marcos
Post by Kay Gottschalk
Hi there,
is there any way not to run g_rdf interactively? what I mean is the
following: I have many (>100) index files, and I want to run g_rdf with
the first group in the index file as reference as the other for the
distribution function. Doing it >100 times by hand is kind of annoying.
I guess there is some line command which can do it, but somehow I
couldn't figure it out.
Thanks for your help,
Kay.
--
Marcos Villarreal
Grupo de Biofisica
Departamento de Quimica Biologica - CIQUIBIC.
Universidad Nacional de Cordoba.
Cordoba. Argentina.
http://www.fcq.unc.edu.ar/ciquibic
Nguyen Hoang Phuong
2003-10-02 20:44:01 UTC
Permalink
Hi, just name the index files as index_1,ndx, index_2.ndx...index_N.ndx

then write the script file:


c=0
while [ $c -le N ]; do
let c=c+1

g_rdf -f *.trr -s *tpr -n index_${c}.ndx -o rdf_${c}

done


Hope it helps.

Phuong
Post by Kay Gottschalk
Hi there,
is there any way not to run g_rdf interactively? what I mean is the
following: I have many (>100) index files, and I want to run g_rdf with
the first group in the index file as reference as the other for the
distribution function. Doing it >100 times by hand is kind of annoying.
I guess there is some line command which can do it, but somehow I
couldn't figure it out.
Thanks for your help,
Kay.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Kay Gottschalk
2003-10-02 20:57:01 UTC
Permalink
thanks, I'll try...
K.
Post by Marcos Villarreal
you can do it with a bash script
#!/bin/bash
n=5 #number of rdf to calculate
g_rdf -n -s <<+
$n
$first #the reference group
1
2
3
4
5
+
I hope this helps.
Saludos,
Marcos
Post by Kay Gottschalk
Hi there,
is there any way not to run g_rdf interactively? what I mean is the
following: I have many (>100) index files, and I want to run g_rdf
with
the first group in the index file as reference as the other for the
distribution function. Doing it >100 times by hand is kind of
annoying.
I guess there is some line command which can do it, but somehow I
couldn't figure it out.
Thanks for your help,
Kay.
--
Marcos Villarreal
Grupo de Biofisica
Departamento de Quimica Biologica - CIQUIBIC.
Universidad Nacional de Cordoba.
Cordoba. Argentina.
http://www.fcq.unc.edu.ar/ciquibic
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Ilya Chorny
2003-10-03 00:12:01 UTC
Permalink
Hello

In the manual it says that by default sc_sigma = .3 nm.
What C6 value does this correspond to. More specifically I have C6 set
to 0.0. What will gromacs use as the C6 value in the soft core
potential.

Thanks

Ilya
--
----------------------------------------------------------------
Ilya Chorny Ph.D. Department of Pharmaceutical Chemistry
Postdoctoral Researcher University of California-San Francisco
phone: (408)-887-8496 Genentech Hall 600 16th St Box 2240
fax: (415)502-1411 San Francisco, Ca, 94107
email: ichorny at maxwell.ucsf.edu
----------------------------------------------------------------
Ilya Chorny
2003-10-07 02:15:04 UTC
Permalink
Hi I was wondering if anybody has an answer to my question?

Ilya
Post by Ilya Chorny
Hello
In the manual it says that by default sc_sigma = .3 nm.
What C6 value does this correspond to. More specifically I have C6 set
to 0.0. What will gromacs use as the C6 value in the soft core
potential.
Thanks
Ilya
Erik Lindahl
2003-10-07 02:51:01 UTC
Permalink
Hi,

The softcore sigma value is used for the softcore distortion/scaling of
the potential, not the actual interaction.

If both c6 & c12 are positive, it is calculated from sigma^6=c12/c6,
otherwise the default value is used.


Hence, If you have c6=0 for a particular interaction, the dispersion
part will be zero no matter what the sc_sigma value is.

Cheers,

Erik
Post by Ilya Chorny
Hi I was wondering if anybody has an answer to my question?
Ilya
Post by Ilya Chorny
Hello
In the manual it says that by default sc_sigma = .3 nm.
What C6 value does this correspond to. More specifically I have C6 set
to 0.0. What will gromacs use as the C6 value in the soft core
potential.
Thanks
Ilya
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Ilya Chorny
2003-10-07 03:13:01 UTC
Permalink
I don't understand. I'm am only interested in what happens to the
softcore distortion if C6 = 0. In this case is sigma^6 = sc_sigma^6.

Thanks

Ilya
Post by Erik Lindahl
Hi,
The softcore sigma value is used for the softcore distortion/scaling of
the potential, not the actual interaction.
If both c6 & c12 are positive, it is calculated from sigma^6=c12/c6,
otherwise the default value is used.
Hence, If you have c6=0 for a particular interaction, the dispersion
part will be zero no matter what the sc_sigma value is.
Cheers,
Erik
Post by Ilya Chorny
Hi I was wondering if anybody has an answer to my question?
Ilya
Post by Ilya Chorny
Hello
In the manual it says that by default sc_sigma = .3 nm.
What C6 value does this correspond to. More specifically I have C6 set
to 0.0. What will gromacs use as the C6 value in the soft core
potential.
Thanks
Ilya
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Erik Lindahl
2003-10-07 03:41:01 UTC
Permalink
Hi,

As a said in the last paragraph - if c6=0 there isn't any dispersion
interaction at all, neither distorted or normal.

Just go back to the definition in section 4.3.1 of the manual.
Interactions are additive, so the formula holds true for the individual
dispersion and repulsion parts of the Lennard-Jones potential.

Cheers,

Erik
Post by Ilya Chorny
I don't understand. I'm am only interested in what happens to the
softcore distortion if C6 = 0. In this case is sigma^6 = sc_sigma^6.
Thanks
Ilya
Post by Erik Lindahl
Hi,
The softcore sigma value is used for the softcore distortion/scaling
of
the potential, not the actual interaction.
If both c6 & c12 are positive, it is calculated from sigma^6=c12/c6,
otherwise the default value is used.
Hence, If you have c6=0 for a particular interaction, the dispersion
part will be zero no matter what the sc_sigma value is.
Cheers,
Erik
Post by Ilya Chorny
Hi I was wondering if anybody has an answer to my question?
Ilya
Post by Ilya Chorny
Hello
In the manual it says that by default sc_sigma = .3 nm.
What C6 value does this correspond to. More specifically I have C6
set
to 0.0. What will gromacs use as the C6 value in the soft core
potential.
Thanks
Ilya
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Ilya Chorny
2003-10-07 04:04:00 UTC
Permalink
More specifically. My C12 term has a positive value. My C6 term is 0. I
am only concerned with the C12 term which is
C12/(alpha*sigma^6*lambda^2+r^6)^2
but the sigma value must be C12/C6 which if C6 was zero would blow up.
Does gromacs set the C6 value to sc_sigma^6.

Ilya


If the soft core used a value of zero then it would blow up. What
Post by Erik Lindahl
Hi,
As a said in the last paragraph - if c6=0 there isn't any dispersion
interaction at all, neither distorted or normal.
Just go back to the definition in section 4.3.1 of the manual.
Interactions are additive, so the formula holds true for the individual
dispersion and repulsion parts of the Lennard-Jones potential.
Cheers,
Erik
Post by Ilya Chorny
I don't understand. I'm am only interested in what happens to the
softcore distortion if C6 = 0. In this case is sigma^6 = sc_sigma^6.
Thanks
Ilya
Post by Erik Lindahl
Hi,
The softcore sigma value is used for the softcore distortion/scaling
of
the potential, not the actual interaction.
If both c6 & c12 are positive, it is calculated from sigma^6=c12/c6,
otherwise the default value is used.
Hence, If you have c6=0 for a particular interaction, the dispersion
part will be zero no matter what the sc_sigma value is.
Cheers,
Erik
Post by Ilya Chorny
Hi I was wondering if anybody has an answer to my question?
Ilya
Post by Ilya Chorny
Hello
In the manual it says that by default sc_sigma = .3 nm.
What C6 value does this correspond to. More specifically I have C6
set
to 0.0. What will gromacs use as the C6 value in the soft core
potential.
Thanks
Ilya
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Erik Lindahl
2003-10-07 05:14:00 UTC
Permalink
Hi,

No - the sigma value will be the default sc_sigma since c6 was zero.

Actually, the sigma in the softcore equations is just an interaction
lengthscale that could be completely separate from the actual LJ
interaction parameters c6/c12. In practice it is convenient to
calculate it from the LJ parameters if they are availabe, so that's
what everybody does unless c6 or c12 is zero.

Think of it this way:

1. First we calculate or assign the length sigma. In most cases it is
smart to use c6/c12 to derive a reasonable length, but it doesn't
really relate to the LJ interaction. If we cannot derive it, use
sigma=sc_sigma.
2. Forget that this value of sigma might have been derived from c6/c12
- it is only a rough approximation anyway.
3. Use the calculated value of sigma for the soft-core distortion of
potentials, including LJ interactions if c6 or c12 is non-zero.

If c12>0 and c6=0, the default sc_sigma will be used to distort the
repulsive potential. The dispersion interaction is zero with c6=0.

Cheers,

Erik
Post by Ilya Chorny
More specifically. My C12 term has a positive value. My C6 term is 0. I
am only concerned with the C12 term which is
C12/(alpha*sigma^6*lambda^2+r^6)^2
but the sigma value must be C12/C6 which if C6 was zero would blow up.
Does gromacs set the C6 value to sc_sigma^6.
Ilya
If the soft core used a value of zero then it would blow up. What
Post by Erik Lindahl
Hi,
As a said in the last paragraph - if c6=0 there isn't any dispersion
interaction at all, neither distorted or normal.
Just go back to the definition in section 4.3.1 of the manual.
Interactions are additive, so the formula holds true for the
individual
dispersion and repulsion parts of the Lennard-Jones potential.
Cheers,
Erik
Post by Ilya Chorny
I don't understand. I'm am only interested in what happens to the
softcore distortion if C6 = 0. In this case is sigma^6 = sc_sigma^6.
Thanks
Ilya
Post by Erik Lindahl
Hi,
The softcore sigma value is used for the softcore distortion/scaling
of
the potential, not the actual interaction.
If both c6 & c12 are positive, it is calculated from sigma^6=c12/c6,
otherwise the default value is used.
Hence, If you have c6=0 for a particular interaction, the dispersion
part will be zero no matter what the sc_sigma value is.
Cheers,
Erik
Post by Ilya Chorny
Hi I was wondering if anybody has an answer to my question?
Ilya
Post by Ilya Chorny
Hello
In the manual it says that by default sc_sigma = .3 nm.
What C6 value does this correspond to. More specifically I have C6
set
to 0.0. What will gromacs use as the C6 value in the soft core
potential.
Thanks
Ilya
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
_______________________________________________
gmx-users mailing list
gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Kay Gottschalk
2003-10-15 19:08:01 UTC
Permalink
Hi Guys amd Gals,

I'd like to do a minor modification on g_rdf: if I provide an index
file consisting just of two groups, I'd like it to take automatically
the first group as reference group without asking further questions. I
have the feeling that this should be rather easy - but stupidly enough
my programing skils a comparable to those of a drosophila. Therefore,
I'd be more than grateful for advise on that.
Thanks,
Kay.
Marc Baaden
2003-10-15 19:28:01 UTC
Permalink
Hi Kay,

why change the code for this ?
You can always use an

echo -e "0\n" | g_rdf ...

which does the same thing (IMHO).

Cheers,
Marc
Post by Kay Gottschalk
I'd like to do a minor modification on g_rdf: if I provide an index
file consisting just of two groups, I'd like it to take automatically
the first group as reference group without asking further questions. I
have the feeling that this should be rather easy - but stupidly enough
my programing skils a comparable to those of a drosophila. Therefore,
I'd be more than grateful for advise on that.
--
Dr. Marc Baaden - Institut de Biologie Physico-Chimique, Paris
mailto:baaden at smplinux.de - http://www.marc-baaden.de
FAX: +49 697912 39550 - Tel: +33 15841 5176 ou +33 609 843217
Kay Gottschalk
2003-10-15 20:26:01 UTC
Permalink
Hey Marc,

you are my god, may I worship you?
Thanks a lot,
K.

Continue reading on narkive:
Loading...