Re: [gmx-users] Re: Periodic Boundary Conditions g_mindist -pi

2011-02-16 Thread Tsjerk Wassenaar 
On Wed, Feb 16, 2011 at 1:00 PM, Mark Abraham  wrote:
> On 16/02/2011 10:30 PM, ifat shub wrote:
>
> Hi Tsjerk,
> Thank you for your reply.
> I am aware of the trajconv option but I wanted to know if there is a way to
> avoid these kind of jumps over the periodic boundaries during the mdrun and
> not post process?
>
> No. mdrun does not know in advance what your visualization requirements are,
> and frankly there are better things to do with expensive compute cluster
> time. Post-processing a small number of frames elsewhere is much better use
> of resources.
>

Or:

No. mdrun takes the internal representation that is most efficient,
and there are better things to do with expensive compute cluster time
than writing out a trajectory without jumps, which might not even be
what you want in the end. :)

Tsjerk

> Mark
>
>
>> message: 4
>> Date: Wed, 16 Feb 2011 11:19:14 +0200
>> From: ifat shub 
>> Subject: [gmx-users] Periodic Boundary Conditions g_mindist -pi
>> To: gmx-users@gromacs.org
>> Message-ID:
>>        
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> Hi,
>>
>>
>>
>> I am running a simulation on the complex 1aik.pdb in 310K. I wanted to see
>> if the complex is seeing its next periodic image, so I used the g_mindist
>> command with the -pi option. My command line was:
>>
>> g_mindist -f run.xtc -s run.gro -n index.ndx -od tmp.xvg -pi
>>
>> The output (see below) was stable until ~344ps when there is a  jump in
>> the
>> max internal distance (third column) from ~6nm to ~22nm. After the jump
>> the
>> numbers are reduced back to ~6nm and remained stable until the run is
>> completed at 1ns.
>>
>> Does anyone know how to explain this jump? Is this a real problem or just
>> a
>> visualization artifact? Is there a way to avoid such jumps?
>>
>>
>>
>> Here is the mdp file I used:
>>
>> --run.mdp--
>>
>> integrator      = md
>>
>> nsteps          = 100
>>
>> dt              = 0.001
>>
>> coulombtype     = pme
>>
>> vdw-type        = cut-off
>>
>> tcoupl          = Berendsen
>>
>> tc-grps         = protein non-protein
>>
>> tau-t           = 0.1 0.1
>>
>> ref-t           = 310 310
>>
>> nstxout         = 100
>>
>> nstvout         = 0
>>
>> nstxtcout       = 100
>>
>> nstenergy           = 100
>>
>> comm_mode     = Linear ; Angular
>>
>> comm_grps        = Protein
>>
>> xtc_grps               = Protein
>>
>> energygrps         = Protein
>>
>> --
>>
>>
>>
>> Thanks,
>>
>> Ifat
>>
>>
>>
>> The output:
>>
>> 343.7   10.813 5.924   16.445 16.445 16.445
>>
>> 343.8   10.809 5.949   16.445 16.445 16.445
>>
>> 343.9   10.804 5.959   16.445 16.445 16.445
>>
>> 344      10.808 5.974   16.445 16.445 16.445
>>
>> 344.1   0.18     21.982 16.445 16.445 16.445
>>
>> 344.2   10.778 5.977   16.445 16.445 16.445
>>
>> 344.3   10.768 5.996   16.445 16.445 16.445
>>
>> 344.4   10.764 6.016   16.445 16.445 16.445
>>
>> 344.5   10.722 6.029   16.445 16.445 16.445
>>
>> 344.6   10.774 6.01     16.445 16.445 16.445
>>
>> 344.7   0.174   21.984 16.445 16.445 16.445
>>
>> 344.8   0.176   21.98   16.445 16.445 16.445
>>
>> 344.9   0.17     22.002 16.445 16.445 16.445
>>
>> 345      0.173   21.981 16.445 16.445 16.445
>>
>> 345.1   0.191   21.954 16.445 16.445 16.445
>>
>> 345.2   0.183   21.958 16.445 16.445 16.445
>>
>> 345.3   0.181   22.012 16.445 16.445 16.445
>>
>> 345.4   0.17     22.054 16.445 16.445 16.445
>>
>> 345.5   0.168   22.054 16.445 16.445 16.445
>>
>> 345.6   0.189   22.039 16.445 16.445 16.445
>>
>> 345.7   0.171   22.007 16.445 16.445 16.445
>>
>> 345.8   0.186   22.031 16.445 16.445 16.445
>>
>> 345.9   0.171   22.077 16.445 16.445 16.445
>>
>> 346      0.187   21.99   16.445 16.445 16.445
>>
>> 346.1   0.173   21.984 16.445 16.445 16.445
>>
>> 346.2   0.181   22.02   16.445 16.445 16.445
>>
>> 346.3   10.82   5.984   16.445 16.445 16.445
>>
>> 346.4   10.81   6.002   16.445 16.445 16.445
>>
>> 346.5   10.819 6.008   16.445 16.445 16.445
>>
>> 346.6   10.813 5.996   16.445 16.445 16.445
>>
>> 346.7   10.781 6.006   16.445 16.445 16.445
>>
>> 346.8   10.793 6.026   16.445 16.445 16.445
>>
>> 346.9   10.745 5.985   16.445 16.445 16.445
>>
>> 347      10.762 5.999   16.445 16.445 16.445
>>
>> 347.1   10.781 5.984   16.445 16.445 16.445
>>
>> 347.2   10.784 6.002   16.445 16.445 16.445
>> -- next part --
>> An HTML attachment was scrubbed...
>> URL:
>> http://lists.gromacs.org/pipermail/gmx-users/attachments/20110216/8495d798/attachment-0001.html
>>
>> --
>>
>> Message: 5
>> Date: Wed, 16 Feb 2011 10:43:56 +0100
>> From: Tsjerk Wassenaar 
>> Subject: Re: [gmx-users] Periodic Boundary Conditions g_mindist -pi
>> To: Discussion list for GROMACS users 
>> Message-ID:
>>        
>> Content-Type: text/plain; charset=ISO-8859-1
>>
>> Hi Ifat,
>>
>> I guess this is a jump over the periodic boundaries. You should remove
>> jumps from the trajectory (-pbc nojump) before running g_mindist -pi.
>>
>> Cheers,
>>
>> Tsjerk
>>
>> On Wed, Feb 1

Re: [gmx-users] Re: Periodic Boundary Conditions g_mindist -pi

2011-02-16 Thread Mark Abraham 

On 16/02/2011 10:30 PM, ifat shub wrote:


Hi Tsjerk,
Thank you for your reply.
I am aware of the trajconv option but I wanted to know if there is a 
way to avoid these kind of jumps over the periodic boundaries during 
the mdrun and not post process?


No. mdrun does not know in advance what your visualization requirements 
are, and frankly there are better things to do with expensive compute 
cluster time. Post-processing a small number of frames elsewhere is much 
better use of resources.


Mark



message: 4
Date: Wed, 16 Feb 2011 11:19:14 +0200
From: ifat shub mailto:shubi...@gmail.com>>
Subject: [gmx-users] Periodic Boundary Conditions g_mindist -pi
To: gmx-users@gromacs.org 
Message-ID:
mailto:sgjftmrf-0nvmzzogfs%2bxhyxv5u6ggfnhr...@mail.gmail.com>>
Content-Type: text/plain; charset="iso-8859-1"

Hi,



I am running a simulation on the complex 1aik.pdb in 310K. I
wanted to see
if the complex is seeing its next periodic image, so I used the
g_mindist
command with the -pi option. My command line was:

g_mindist -f run.xtc -s run.gro -n index.ndx -od tmp.xvg -pi

The output (see below) was stable until ~344ps when there is a
 jump in the
max internal distance (third column) from ~6nm to ~22nm. After the
jump the
numbers are reduced back to ~6nm and remained stable until the run is
completed at 1ns.

Does anyone know how to explain this jump? Is this a real problem
or just a
visualization artifact? Is there a way to avoid such jumps?



Here is the mdp file I used:

--run.mdp--

integrator  = md

nsteps  = 100

dt  = 0.001

coulombtype = pme

vdw-type= cut-off

tcoupl  = Berendsen

tc-grps = protein non-protein

tau-t   = 0.1 0.1

ref-t   = 310 310

nstxout = 100

nstvout = 0

nstxtcout   = 100

nstenergy   = 100

comm_mode = Linear ; Angular

comm_grps= Protein

xtc_grps   = Protein

energygrps = Protein

--



Thanks,

Ifat



The output:

343.7   10.813 5.924   16.445 16.445 16.445

343.8   10.809 5.949   16.445 16.445 16.445

343.9   10.804 5.959   16.445 16.445 16.445

344  10.808 5.974   16.445 16.445 16.445

344.1   0.18 21.982 16.445 16.445 16.445

344.2   10.778 5.977   16.445 16.445 16.445

344.3   10.768 5.996   16.445 16.445 16.445

344.4   10.764 6.016   16.445 16.445 16.445

344.5   10.722 6.029   16.445 16.445 16.445

344.6   10.774 6.01 16.445 16.445 16.445

344.7   0.174   21.984 16.445 16.445 16.445

344.8   0.176   21.98   16.445 16.445 16.445

344.9   0.17 22.002 16.445 16.445 16.445

345  0.173   21.981 16.445 16.445 16.445

345.1   0.191   21.954 16.445 16.445 16.445

345.2   0.183   21.958 16.445 16.445 16.445

345.3   0.181   22.012 16.445 16.445 16.445

345.4   0.17 22.054 16.445 16.445 16.445

345.5   0.168   22.054 16.445 16.445 16.445

345.6   0.189   22.039 16.445 16.445 16.445

345.7   0.171   22.007 16.445 16.445 16.445

345.8   0.186   22.031 16.445 16.445 16.445

345.9   0.171   22.077 16.445 16.445 16.445

346  0.187   21.99   16.445 16.445 16.445

346.1   0.173   21.984 16.445 16.445 16.445

346.2   0.181   22.02   16.445 16.445 16.445

346.3   10.82   5.984   16.445 16.445 16.445

346.4   10.81   6.002   16.445 16.445 16.445

346.5   10.819 6.008   16.445 16.445 16.445

346.6   10.813 5.996   16.445 16.445 16.445

346.7   10.781 6.006   16.445 16.445 16.445

346.8   10.793 6.026   16.445 16.445 16.445

346.9   10.745 5.985   16.445 16.445 16.445

347  10.762 5.999   16.445 16.445 16.445

347.1   10.781 5.984   16.445 16.445 16.445

347.2   10.784 6.002   16.445 16.445 16.445
-- next part --
An HTML attachment was scrubbed...
URL:

http://lists.gromacs.org/pipermail/gmx-users/attachments/20110216/8495d798/attachment-0001.html

--

Message: 5
Date: Wed, 16 Feb 2011 10:43:56 +0100
From: Tsjerk Wassenaar mailto:tsje...@gmail.com>>
Subject: Re: [gmx-users] Periodic Boundary Conditions g_mindist -pi
To: Discussion list for GROMACS users mailto:gmx-users@gromacs.org>>
Message-ID:
mailto:abxfur%2badkenmxem...@mail.gmail.com>>
Content-Type: text/plain; charset=ISO-8859-1

Hi Ifat,

I guess this is a jump over the periodic boundaries. You should remove
jumps from the trajectory (-pbc nojump) before running g_mindist -pi.

Cheers,

Tsjerk

On Wed, Feb 16, 2011 at 10:19 AM, ifat shub mailto:shubi...@gmail.com>> wrote:
> Hi,
>
>
>
> I am running a simulation on the complex 1aik.pdb in 310K. I

[gmx-users] Re: Periodic Boundary Conditions g_mindist -pi

2011-02-16 Thread ifat shub 
Hi Tsjerk,
Thank you for your reply.
I am aware of the trajconv option but I wanted to know if there is a way to
avoid these kind of jumps over the periodic boundaries during the mdrun and
not post process?
Thanks,
Ifat

 message: 4
> Date: Wed, 16 Feb 2011 11:19:14 +0200
> From: ifat shub 
> Subject: [gmx-users] Periodic Boundary Conditions g_mindist -pi
> To: gmx-users@gromacs.org
> Message-ID:
>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Hi,
>
>
>
> I am running a simulation on the complex 1aik.pdb in 310K. I wanted to see
> if the complex is seeing its next periodic image, so I used the g_mindist
> command with the -pi option. My command line was:
>
> g_mindist -f run.xtc -s run.gro -n index.ndx -od tmp.xvg -pi
>
> The output (see below) was stable until ~344ps when there is a  jump in the
> max internal distance (third column) from ~6nm to ~22nm. After the jump the
> numbers are reduced back to ~6nm and remained stable until the run is
> completed at 1ns.
>
> Does anyone know how to explain this jump? Is this a real problem or just a
> visualization artifact? Is there a way to avoid such jumps?
>
>
>
> Here is the mdp file I used:
>
> --run.mdp--
>
> integrator  = md
>
> nsteps  = 100
>
> dt  = 0.001
>
> coulombtype = pme
>
> vdw-type= cut-off
>
> tcoupl  = Berendsen
>
> tc-grps = protein non-protein
>
> tau-t   = 0.1 0.1
>
> ref-t   = 310 310
>
> nstxout = 100
>
> nstvout = 0
>
> nstxtcout   = 100
>
> nstenergy   = 100
>
> comm_mode = Linear ; Angular
>
> comm_grps= Protein
>
> xtc_grps   = Protein
>
> energygrps = Protein
>
> --
>
>
>
> Thanks,
>
> Ifat
>
>
>
> The output:
>
> 343.7   10.813 5.924   16.445 16.445 16.445
>
> 343.8   10.809 5.949   16.445 16.445 16.445
>
> 343.9   10.804 5.959   16.445 16.445 16.445
>
> 344  10.808 5.974   16.445 16.445 16.445
>
> 344.1   0.18 21.982 16.445 16.445 16.445
>
> 344.2   10.778 5.977   16.445 16.445 16.445
>
> 344.3   10.768 5.996   16.445 16.445 16.445
>
> 344.4   10.764 6.016   16.445 16.445 16.445
>
> 344.5   10.722 6.029   16.445 16.445 16.445
>
> 344.6   10.774 6.01 16.445 16.445 16.445
>
> 344.7   0.174   21.984 16.445 16.445 16.445
>
> 344.8   0.176   21.98   16.445 16.445 16.445
>
> 344.9   0.17 22.002 16.445 16.445 16.445
>
> 345  0.173   21.981 16.445 16.445 16.445
>
> 345.1   0.191   21.954 16.445 16.445 16.445
>
> 345.2   0.183   21.958 16.445 16.445 16.445
>
> 345.3   0.181   22.012 16.445 16.445 16.445
>
> 345.4   0.17 22.054 16.445 16.445 16.445
>
> 345.5   0.168   22.054 16.445 16.445 16.445
>
> 345.6   0.189   22.039 16.445 16.445 16.445
>
> 345.7   0.171   22.007 16.445 16.445 16.445
>
> 345.8   0.186   22.031 16.445 16.445 16.445
>
> 345.9   0.171   22.077 16.445 16.445 16.445
>
> 346  0.187   21.99   16.445 16.445 16.445
>
> 346.1   0.173   21.984 16.445 16.445 16.445
>
> 346.2   0.181   22.02   16.445 16.445 16.445
>
> 346.3   10.82   5.984   16.445 16.445 16.445
>
> 346.4   10.81   6.002   16.445 16.445 16.445
>
> 346.5   10.819 6.008   16.445 16.445 16.445
>
> 346.6   10.813 5.996   16.445 16.445 16.445
>
> 346.7   10.781 6.006   16.445 16.445 16.445
>
> 346.8   10.793 6.026   16.445 16.445 16.445
>
> 346.9   10.745 5.985   16.445 16.445 16.445
>
> 347  10.762 5.999   16.445 16.445 16.445
>
> 347.1   10.781 5.984   16.445 16.445 16.445
>
> 347.2   10.784 6.002   16.445 16.445 16.445
> -- next part --
> An HTML attachment was scrubbed...
> URL:
> http://lists.gromacs.org/pipermail/gmx-users/attachments/20110216/8495d798/attachment-0001.html
>
> --
>
> Message: 5
> Date: Wed, 16 Feb 2011 10:43:56 +0100
> From: Tsjerk Wassenaar 
> Subject: Re: [gmx-users] Periodic Boundary Conditions g_mindist -pi
> To: Discussion list for GROMACS users 
> Message-ID:
>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Hi Ifat,
>
> I guess this is a jump over the periodic boundaries. You should remove
> jumps from the trajectory (-pbc nojump) before running g_mindist -pi.
>
> Cheers,
>
> Tsjerk
>
> On Wed, Feb 16, 2011 at 10:19 AM, ifat shub  wrote:
> > Hi,
> >
> >
> >
> > I am running a simulation on the complex 1aik.pdb in 310K. I wanted to
> see
> > if the complex is seeing its next periodic image, so I used the g_mindist
> > command with the -pi option. My command line was:
> >
> > g_mindist -f run.xtc -s run.gro -n index.ndx -od tmp.xvg -pi
> >
> > The output (see below) was stable until ~344ps when there is a  jump in
> the
> > max internal distance (third column) from ~6nm to ~22nm. After the jump
> the
> > numbers are reduced back to ~6nm and remained stable until the run is
> > completed at 1ns.
> >
> > Does anyone know how to explain this jump? Is this a real problem or just
> a
> > visualization artifact? Is there a way to avoid such jumps?
> >
> >
> >
> > Here is the mdp file I u