protomol-users — Protomol Developer and User DIscussion List

[Protomol-users] Minimal image output

[Thierry M. <Thi...@bc...>]

As asked, I added minimal image for all output object writing positions. 
The approach is the same as with the output frequency, a global flag 
defining the default behavior and a local flag for each output object, 
which can be specified individually, if needed. By default ProtoMol will 
write a minimal image. Furthermore, the minimal image of positions will 
not break the property of correct intra-molecular distances, meaning 
molecules are not wrapped partially.

-Thierry

Re: [Protomol-users] Re: Compiling ProtoMol with gcc4.0 (fwd)

[Thierry M. <Thi...@bc...>]

Dan,

I changed the unamed enumerator to AngleType and changed also the its data 
member type. It should fix your problem. I did also some test with GCC 4.0.1, 
but did not encounter the problem ... anyway it's nicer/better to use it named 
and use the type. Hopes that it helps ..

Bests,
-Thierry

Re: [Protomol-users] Re: Compiling ProtoMol with gcc4.0

[Thierry M. <Thi...@bc...>]

Dan,

I changed the unamed enumerator to AngleType and changed also the its data 
member type. It should fix your problem. I did also some test with GCC 
4.0.1, but did not encounter the problem ... anyway it's nicer/better to 
use it named and use the type. Hopes that it helps ..

Bests,
-Thierry

[Protomol-users] Compiling ProtoMol with gcc4.0

[Dan K. <dw...@ma...>]

For anyone running Fedora Core 4 or is using gcc4 here is my edit to 
compile protomol:
gcc4.0 apparently  doesn't like unamed enums:

./framework/topology/AngleInfo.h  line 17.

CHANGE:
 
enum { ANGLE_VALUE  =    10001,
           ANGLE_POINTER =   10002,
           ANGLE_NOTSET =    10003,
           VISITED     =     10001,
           NOT_VISITED =     10002};

TO:
 enum angle_state { ANGLE_VALUE  =    10001,
                              ANGLE_POINTER =   10002,
                              ANGLE_NOTSET =    10003,
                              VISITED     =     10001,
                              NOT_VISITED =     10002};


I've run a few of the examples, and this quick fix has not broke 
anything yet.
 
Version: protmol-2.0.3
System:  FC4, dual 3 Ghz P4, 512 RAM,gcc  4.0.1

I'm not sure where else to submit this to, so I figured this list would 
be the best place.  The real solution should have a GCC version check, 
if for some reason it needs to be an anonymous enum.

~Dan Kulp

Re: [Protomol-users] Caching facilities of output data

[Paul R B. <pbr...@nd...>]

Thierry,

Some of the outputs (such as XYZVelFile) only inherit from the Output.h object
and not OutputFile.h and thus did not benefit from the caching.  Is there a
design reason why all of the output modules which write to file do not inherit
from OutputFile?  I am using both output types and would like to reduce the
filesystem access overhead (while still getting the data resolution I need :)

Paul R Brenner, P.E.
Graduate Research Assistant
Computer Science and Engineering
The University of Notre Dame


Quoting Thierry Matthey <Thi...@bc...>:

> Hi,
>
> I revisited OutputFile, which already provided an interface to write
> and append data to a file. I extended it such that it supports caching:
>   - by frequency
>   - by size
> Furthermore, I added an additional keyword such that you can impose a
> minimal time between two writes to close the file temporarily.
>
> What changed:
> - implement doRunCached() for doRun()
> - use myBuffer for myFile
> - OutputFile has a default doFinialize() implementation.
> - OutputFile has a implementation of getParameters, which must be called
>    if you like to add additional keywords.
>
> Cheers,
> Thierry
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: NEC IT Guy Games.  How far can you shotput
> a projector? How fast can you ride your desk chair down the office luge
> track?
> If you want to score the big prize, get to know the little guy.
> Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
> _______________________________________________
> Protomol-users mailing list
> Pro...@li...
> https://lists.sourceforge.net/lists/listinfo/protomol-users
>

[Protomol-users] Re: Compiling ProtoMol with gcc4.0

[Paul R B. <pbr...@nd...>]

Thanks for the insight Dan,

We are trying a couple of tests with the new gcc4.0.x release.  As you indicate
there are a few things this new version of the compiler doesn't like.  We will
probably continue to incorporate the changes up through 4.1.ish and then
consider a releasing a new source tarball.  We are make updates in the cvs
repository and the next source release will likely have builtin functionality
for the Replica Exchange Method using MPI.

Regards,
Paul

Quoting Dan Kulp <dw...@ma...>:

> For anyone running Fedora Core 4 or is using gcc4 here is my edit to
> compile protomol:
> gcc4.0 apparently  doesn't like unamed enums:
>
> ./framework/topology/AngleInfo.h  line 17.
>
> CHANGE:
>
> enum { ANGLE_VALUE  =    10001,
>            ANGLE_POINTER =   10002,
>            ANGLE_NOTSET =    10003,
>            VISITED     =     10001,
>            NOT_VISITED =     10002};
>
> TO:
>  enum angle_state { ANGLE_VALUE  =    10001,
>                               ANGLE_POINTER =   10002,
>                               ANGLE_NOTSET =    10003,
>                               VISITED     =     10001,
>                               NOT_VISITED =     10002};
>
>
> I've run a few of the examples, and this quick fix has not broke
> anything yet.
>
> Version: protmol-2.0.3
> System:  FC4, dual 3 Ghz P4, 512 RAM,gcc  4.0.1
>
> I'm not sure where else to submit this to, so I figured this list would
> be the best place.  The real solution should have a GCC version check,
> if for some reason it needs to be an anonymous enum.
>
> ~Dan Kulp
>
>

Re: [Protomol-users] Caching facilities of output data

[Thierry M. <Thi...@bc...>]

Paul,

> Some of the outputs (such as XYZVelFile) only inherit from the Output.h o=
bject
> and not OutputFile.h and thus did not benefit from the caching.  Is there=
 a
> design reason why all of the output modules which write to file do not in=
herit
> from OutputFile?  I am using both output types and would like to reduce t=
he
> filesystem access overhead (while still getting the data resolution I nee=
d :)

The main idea with OutputFile is to provide a more specialized base class=
=20
(caching etc.) for all output appending some data (scalars: e.g., time,
temperature, volume) to one file at a given output frequency. In case
of XYZVelFile (class OutputXYZTrajectoryVel), each time the velocities
are dumped and appended to the file by XYZTrajectoryWriter; the same
yields for DCD, and other complex output formats, which are appended
to the same file. To enable caching one can either:

1) cache the data in the output class and some how
    pass it to the writer (XYZTrajectoryWriter) ... not really the way
    to go ;-),

2) or you re-implement XYZTrajectoryWriter in the output class ... code
    duplication, fy, fy,

3) or you enable class Writer with caching options, which not
    all will use and make it hard to add new writer classes to the
    io-lib ...

4) or you add a new abstract class WriterCached from which
    writer classes can inherit caching - in the same way as done in
    OutputFile, but how you can get around with ASCII and binaries is
    an other story,

5) or you add explicitly caching the to the writers in question ... code
    duplication, again.

At present, the possible candidates for caching are the trajectory
writers: DCD- and XYZTrajectoryWriter, the other writers write a
complete file --- no appending --- should be optimized specially, inside
the class. In case of trajectories one could try to do some caching, but
I'm not that sure if you gain much, since those writers write much
more than one line at the time ...

In case you like to write some data (scalars) and write a trajectory=20
I recommend to inherit from OutputFile, write the scalers via myBuffer
and add trajectory writer by containment, as a data member. Take a
look at OutputPaulTrap, which writes some scalers and dumps the
positions when ever a lower entry state is found by overwriting the
old one.

> Quoting Thierry Matthey <Thi...@bc...>:
>
>> Hi,
>>
>> I revisited OutputFile, which already provided an interface to write
>> and append data to a file. I extended it such that it supports caching:
>>   - by frequency
>>   - by size
>> Furthermore, I added an additional keyword such that you can impose a
>> minimal time between two writes to close the file temporarily.
>>
>> What changed:
>> - implement doRunCached() for doRun()
>> - use myBuffer for myFile
>> - OutputFile has a default doFinialize() implementation.
>> - OutputFile has a implementation of getParameters, which must be called
>>    if you like to add additional keywords.

-Thierry=20
-------------------------+------------------------------------------
  Dr.Sc. Thierry Matthey  | work +47 55584167  | ma...@bc...
  Para//ab, BCCS          | fax  +47 55584295  | www.matthey.org
  Thorm=F8hlensgt. 55       | home +47 55313146  | ICQ# 23561635
  N-5008 Bergen Norway    | www.parallab.no    | E/G/F/N
-------------------------+------------------------------------------

Re: [Protomol-users] BPTI Demo simulation

[Paul R B. <pbr...@nd...>]

Dan,

I've inserted a different config file for you to try.  To use NVT you will need
to specify an NVT integrator like LangevinImpulse.  I ran a quick test of the
inserted config on the ~1000 atom solvated bpti for 10ps and it seemed quite
stable.  When I reference the ~1,000 atom bpti I'm referencing the files which
can be found in the examples directory in the source code.

It may be that the initial positions file on the examples page of the website is
not sufficiently equilibrated.  I would stick with the examples in the source
code as they are more often exercised.   I will ask Theirrey (one of our
developers) if he has had any trouble with the version on the examples page.

WRT the boundary conditions normal and vacuum are synonomous.  Periodic
boundaries are used with NVT as the constant volume is held for the simulation
box.  For NPT simulations the box size (volume) changes to maintain a constant
pressure.

Regards,
Paul
************ config ********

numsteps 10000
firststep 0

posfile bpti.pdb
psffile bpti.psf
parfile bpti.par
temperature 300

seed 7536031

outputfreq 100

allenergiesfile bpti.out.energies.lf

boundaryConditions Periodic

cellManager Cubic
cellsize 4

Integrator {

    level 0 LangevinImpulse {
        timestep 1
	temperature 300
        gamma 10

        force Improper
        force Dihedral
        force Bond
        force Angle
        force Coulomb
          -algorithm NonbondedCutoff
          -switchingFunction C1
          -cutoff 10.0
        force LennardJones
          -algorithm NonbondedCutoff
          -switchingFunction C2
          -cutoff 10.0
          -switchon 0.1
    }

}

************

Quoting Dan Kulp <dw...@ma...>:

> Paul,
>   Thanks for the quick response.  I downloaded both binary and source,
> but the runs I've done are strictly from the binary I downloaded.  I'm
> running on an AMD 2500, but could run just as easily on some p4e's. Both
> systems have redhat 9.0 running.
>
> I'm using the explicit water simulation from the web page (~ 4460 TIP3
> waters).  I've pasted the conf file from the bpti.conf demo from the
> webpage, except with numsteps changed from 100 to 1000000. I also tried
> the NoseNVTLeapfrog integrator (timestep 1, temperature 300, thermal
> 0.5, bathPos 10) ... I tried a bunch of different combinations of
> thermal and bathPos, but all the simulations were unstable at some point
> (I think 4 ps is as far as I got).
>
> Some further questions:
> When boundaryConditions is set to "normal" , what does this mean?  In
> the quickref.pdf there are only two values "Vacuum" and "Periodic".
> What does NVT mean when you have periodic boundaries, isn't the volume,
> in a sense, infinite ? Are there some periodic boundary simulations
> where the size of the center box is allowed to change ?
>
>
> BTW. I'm running protomol version 2.0.3
>
> Thanks,
> Dan
>
>
>
> temperature 300.0
> firststep 0
>
> numsteps 1000000
>
> cellsize 10.0
>
> outputfreq 1
> seed 9783
>
> posfile bpti.pdb
> psffile bpti.psf
> parfile bpti.par
>
> finXYZposfile bpti.out.pos.xyz
> finXYZvelfile bpti.out.vel.xyz
> dcdfile bpti.out.dcd
> allenergiesfile bpti.out.energy
>
> boundaryConditions Normal
> cellManager Cubic
>
>
> Integrator {
>   level 0 Leapfrog {
>         timestep 1
>     force Improper
>     force Dihedral
>     force Bond
>     force Angle
>     force Coulomb
>           -algorithm NonbondedCutoff
>           -switchingFunction C1
>           -cutoff 10.0
>     force LennardJones
>           -algorithm NonbondedCutoff
>           -switchingFunction C2
>           -cutoff 10.0
>           -switchon 0.1
>   }
> }
>
>

Re: [Protomol-users] BPTI Demo simulation

[Paul R B. <pbr...@nd...>]

Hi Dan,

Did you download the source code tarball or the binary from the sourceforge
website?  Which BPTI example are you using (molecule only, solvated ~1,000, or
solvated 14,000 atom).  Which configuration file are you using?  Always lots of
parameters in the world of molecular dynamics.  It should be stable for 1ns with
an NVT integrator (Langevin or Nose-Hoover).

Regards,

Paul R Brenner, P.E.
Graduate Research Assistant
Computer Science and Engineering
The University of Notre Dame


Quoting Dan Kulp <dw...@ma...>:

> Hello all,
>    I've just downloaded protomol and am a new user.  I'd like to run the
> BPTI demo (from the webpage) for 1ns, but the simulation eventually
> becomes unstable.  I would like to run an nvt simulation.  Has anyone
> acheived this on the BPTI demo ?   Would adding in temporary (10's ps?)
> restraints on the waters/protein for a portion of time be a reasonable
> next step to getting a stable simulation? If so, how do you do this in
> protomol?
>
> Thanks,
> Dan
>
>
>
> -------------------------------------------------------
> SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> from IBM. Find simple to follow Roadmaps, straightforward articles,
> informative Webcasts and more! Get everything you need to get up to
> speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> _______________________________________________
> Protomol-users mailing list
> Pro...@li...
> https://lists.sourceforge.net/lists/listinfo/protomol-users
>

[Protomol-users] BPTI Demo simulation

[Dan K. <dw...@ma...>]

Hello all,
   I've just downloaded protomol and am a new user.  I'd like to run the 
BPTI demo (from the webpage) for 1ns, but the simulation eventually 
becomes unstable.  I would like to run an nvt simulation.  Has anyone 
acheived this on the BPTI demo ?   Would adding in temporary (10's ps?) 
restraints on the waters/protein for a portion of time be a reasonable 
next step to getting a stable simulation? If so, how do you do this in 
protomol?

Thanks,
Dan

Re: [Protomol-users] Question about how to use the protomol

[Paul R B. <pbr...@nd...>]

Luen,

I'm not sure where you are pulling the example files from.  It sounds like you
may have an old alanin example config which is calling for the interactive
molecular dynamics module ( ties into the VMD {visual molecular dynamics
application} ).  Please make sure that you are using the most current Windows
executable and the most current example files.

On the following sourceforge website use the following Windows executable and
source code (the source code documentation includes all of the current examples
and documentation).

http://sourceforge.net/project/showfiles.php?group_id=85222&package_id=88124

Download protomol-2.0.3-tar.gz	2391568  Platform-Independent	.gz
Download protomol-2.0.3-Win32.exe	2211840  i386	.exe (32-bit Windows)

Paul R Brenner, P.E.
Graduate Research Assistant
Computer Science and Engineering
The University of Notre Dame


Quoting fung ka luen <lu...@ho...>:

>
>
>
>
> Hi Paul,
> Good morning!
> Thank you for you help
> I have download the "alanin_66.tar.tar" which have .conf, .par, .pdb and .psf
> files.
> I put them all together with the .exe file.
> I run the protomol.exe(name changed) again.
> The error "Failed to open socket."  happened
> Please see the attachment for detail
> However, it is ok for "Water 423"
> Thank you very very much
> Luen
>
>
>
>
>
>
>
> >From: Paul R Brenner <pbr...@nd...>
> >To: fung ka luen <lu...@ho...>
> >CC: pro...@li...
> >Subject: Re: [Protomol-users] Question about how to use the protomol
> >Date: Wed,  6 Jul 2005 08:12:33 -0500
> >
> >Good morning Luen,
> >
> >From the screenshots you sent it looks like the fatal error you are
> experiencing
> >is that Protomol cannot open file alanin.pdb (2nd to last line in screenshot
> >#2).  This is usually do to the fact that Protomol and the configuration
> file
> >are not executed in the same directory as the necessary input files (pdb,
> psf,
> >par).
> >
> >For an example go to the Protomol homepage on sourceforge:
> >
> >http://protomol.sourceforge.net/
> >
> >and click on screenshots.  In
> this example I placed the executable, config file,
> >and necessary input files in the same directory.  If you have only
> downloaded
> >the executable from sourceforge you should also download the source code in
> >which you will find multiple examples and all of the necessary input files.
> >
> >Regards,
> >
> >Paul R Brenner, P.E.
> >Graduate Research Assistant
> >Computer Science and Engineering
> >The University of Notre Dame
> >
> >
> >Quoting fung ka luen <lu...@ho...>:
> >
> > > Dear Sir,
> > >
> > > I am a RA working in City University of Hong Kong. We try to use the
> ProtoMol
> > > for my research project.
> > >
> > > I have downloaded the ProtoMol 2.0.3 and user-guide from your web-site.
> The
> > > detail as follow:
> > >
> > > Porgram
> version:  2.0.3
> > > File Name: protomol-2.0.3-Win32.exe
> > > User-guide: quickref.pdf
> > > config file: alanin.conf.txt
> > > My computer: P4 1.4G , windows XP
> > >
> > >
> > > I run the program in "Command prompt" like ,
> > >
> > > c:\>protomol-2.0.3-Win32.exe alanin.conf.txt
> > >
> > > The result is "No MPI compilation". Please see attachment for detail
> > >
> > > Is there any file I missed? or some software i needed to install?
> > >
> > > Is it the operation steps OK?
> > >
> > > Could you please help me the solve the problem?
> > >
> > > Many thank you!
> > >
> > > Luen
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> >
> >
> >
>
>

Re: [Protomol-users] Question about how to use the protomol

[Thierry M. <Thi...@bc...>]

Dear Luen,

> The result is "No MPI compilation". Please see attachment for detail

Your binary does not support MPI, i.e., only sequential runs.

> Is there any file I missed? or some software i needed to install?

As I can see so far is that protomol.exe could not locate/open the 
position files.

I would recommend to use the GUI TkProtoMol.pl, it's a bit easier to 
launch than from the command line. There is a read-me to install perl on 
Win32/XP.

Cheers,
Thierry

Re: [Protomol-users] Question about how to use the protomol

[Paul R B. <pbr...@nd...>]

Good morning Luen,

From the screenshots you sent it looks like the fatal error you are experiencing
is that Protomol cannot open file alanin.pdb (2nd to last line in screenshot
#2).  This is usually do to the fact that Protomol and the configuration file
are not executed in the same directory as the necessary input files (pdb, psf,
par).

For an example go to the Protomol homepage on sourceforge:

http://protomol.sourceforge.net/

and click on screenshots.  In this example I placed the executable, config file,
and necessary input files in the same directory.  If you have only downloaded
the executable from sourceforge you should also download the source code in
which you will find multiple examples and all of the necessary input files.

Regards,

Paul R Brenner, P.E.
Graduate Research Assistant
Computer Science and Engineering
The University of Notre Dame


Quoting fung ka luen <lu...@ho...>:

> Dear Sir,
>
> I am a RA working in City University of Hong Kong. We try to use the ProtoMol
> for my research project.
>
> I have downloaded the ProtoMol 2.0.3 and user-guide from your web-site. The
> detail as follow:
>
> Porgram version:  2.0.3
> File Name: protomol-2.0.3-Win32.exe
> User-guide: quickref.pdf
> config file: alanin.conf.txt
> My computer: P4 1.4G , windows XP
>
>
> I run the program in "Command prompt" like ,
>
> c:\>protomol-2.0.3-Win32.exe alanin.conf.txt
>
> The result is "No MPI compilation". Please see attachment for detail
>
> Is there any file I missed? or some software i needed to install?
>
> Is it the operation steps OK?
>
> Could you please help me the solve the problem?
>
> Many thank you!
>
> Luen
>
>
>
>
>
>
>
>
>

[Protomol-users] Question about how to use the protomol

[fung ka l. <lu...@ho...>]

<html><div style='background-color:'><DIV class=RTE>Dear Sir,</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>I am a RA working in City University of Hong Kong. We try to use the ProtoMol for my research project.</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>I have downloaded&nbsp;the ProtoMol 2.0.3 and&nbsp;user-guide from your web-site. The detail as follow:</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>Porgram version: &nbsp;2.0.3</DIV>
<DIV class=RTE>File Name: protomol-2.0.3-Win32.exe</DIV>
<DIV class=RTE>User-guide: quickref.pdf</DIV>
<DIV class=RTE>config file: alanin.conf.txt</DIV>
<DIV class=RTE>My computer: P4 1.4G , windows XP</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>I run the program in "Command prompt" like ,</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>c:\&gt;protomol-2.0.3-Win32.exe alanin.conf.txt</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>The result is "No MPI compilation". Please see attachment for detail</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>Is there any file I missed? or some software i needed to install?</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>Is it the operation steps OK?</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>Could you please help me the solve the problem?</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>Many thank you!</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>Luen</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>&nbsp;</DIV>
<DIV class=RTE>&nbsp;</DIV></div></html>

[Protomol-users] Documentation

[Thierry M. <Thi...@bc...>]

Hi,

I just added the documenation to the SF home page: Quick Reference, 
multi-grid, Ewald & PME and performance guidelines.

The source (tex) can be found in CVS.

Cheers,
Thierry

[Protomol-users] possible inclusion of protomol into SPEC MPI2005 benchmark?

[Ron L. <li...@rs...>]

Hello protomol developers

We at the SPEC HPG are developing a MPI set of benchmarks SPEC MPI2005.
We are looking for candidate codes that are parallelized using MPI and
are reasonably scalable from 8-128 ranks. 

Would you consider your code to be a possible candidate for inclusion in
our developing benchmark?

Would you think it would meet the criteria to run for aprox 1 hour across 
16 ranks given an appropriate data set?

Thanks in advance
Ron Lieberman
Vice Chairperson 
SPEC High Performance Group

[Protomol-users] HarmDihedral syntax questions

[Paul R B. <pbr...@nd...>]

Hello Marcelo,

The HarmDihedral force was recently implemented for utilization in Umbrel=
la
Sampling and Coarse Molecular Dynamics techniques.  A sample integrator
specification in the configuration file would be:

# NVT Integrator
Integrator {
  level 0 LangevinImpulse {
    timestep 1.0

    # Target temperature
    temperature 300

    # Parameter
    gamma 5000

    # Bonded forces
    force Bond
    force Angle
    force HarmDihedral
      -kbias 100.0
      -dihedral 0
      -angle 90
      -others true
    force LennardJones
      -algorithm NonbondedSimpleFull
  }
}

To get a listing of parameters for a force or integrator you can type 'pr=
otomol
-f' or 'protomol -i' respectively.  The output of 'protomol -f' for the
HarmDihedral force is:

HarmDihedral
    -kbias                 <real=3D1,non-negative>         # potential bi=
as
constant
    -dihedral              <int=3D0,non-negative>
    -angle                 <real=3D0>      # reference angle -180 to 180
    -others                <boolean=3Dfalse>

which informs us that the reference angle must be specified in angles fro=
m -180
to 180.

With regards to the dihedral index - currently only one dihedral index ca=
n be
idetified, it would not be hard to allow multiple dihedrals but some care=
 would
need to be taken such that consecutive dihedrals do not have conflicting =
biases.
 The index is determined according to the .psf topology file that you gen=
erate.=20
The indices start with 0 and match the order in which they are read from =
the
.psf file.  If you would find the multiple indices option beneficial plea=
se
submit a feature request on the sourceforge site.  I could see how this c=
ould
be beneficial for variations of US however the unbiasing would be challen=
ging.

Hope the weather is nice in Brazil,

Paul R Brenner, P.E.
Graduate Research Assistant
Computer Science and Engineering
The University of Notre Dame


Quoting Marcelo Rossi <mr...@us...>:

> Hi Paul,
>
> I've a new doubt. In HarmDihedral subroutine, how to describe the dihed=
ral
> angles in
> "dihedral" key??
>
>        level 0 LeapFrog {
>                  timestep 0.5
>          force Bond, Angle, Improper
>          force HarmDihedral
>             -algorithm HarmDihedral
>             -dihedral  (???????)
>             -angle 30
>             -others false
>            }
>
> Could be a list of dihedrals angles that i want to study or the complet=
e
> list?? What the
> input format of this dihedrals angles??
>
> Once more, thanks for the attention,
> Regards,
>
> Marcelo.
> --
> Marcelo Rossi
> Faculdade de Medicina
> Universidade de S=E3o Paulo - Brazil
>

[Protomol-users] Caching facilities of output data

[Thierry M. <Thi...@bc...>]

Hi,

I revisited OutputFile, which already provided an interface to write
and append data to a file. I extended it such that it supports caching:
  - by frequency
  - by size
Furthermore, I added an additional keyword such that you can impose a
minimal time between two writes to close the file temporarily.

What changed:
- implement doRunCached() for doRun()
- use myBuffer for myFile
- OutputFile has a default doFinialize() implementation.
- OutputFile has a implementation of getParameters, which must be called
   if you like to add additional keywords.

Cheers,
Thierry

[Protomol-users] DihedralHMC and MDLeapfrog

[Paul R B. <pbr...@nd...>]

Hello Marcelo,

One point of clarification.  I am the author responsible for the Dihedral=
HMC
integrator/propagotor code (It looks like Thierry was think of the Shadow
Hybrid Monte Carlo code when he forwarded you the ppt presentation and Sc=
ott
Hampton's contact info).

I've attached a brief description of the integrator.  An example configur=
ation
file and necessary inputs can be found with the source code in the exampl=
es
directory:

protomol/examples/butane_4/UA_butane.dhmc.conf

The new user's guide for 2.0.3 is in editing.  The users guide for 1.8.3 =
has
most of the the information you need and can be found in the source code
directory:

protomol/doc/UserGuide/

Regards,
Paul

Paul R Brenner, P.E.
Graduate Research Assistant
Computer Science and Engineering
The University of Notre Dame

Quoting Thierry Matthey <Thi...@bc...>:

> Dear Marcelo,
>
> > I would like to have access to the some documentation on Protomol ver=
sion
> 2.0.3, mainly
> > with relation to the DihedralHMC and MDLeapfrog subroutines. Where I =
can
> find them?
> > One more time thanks for the attention,
>
> I forward your request to the developers of the methods:
>
> MDLeapfrog (Ma Qun):
> http://www.cs.njit.edu/~qma/pub/42356.pdf
>
> DihedralHMC (Hampton Scott):
> http://www.nd.edu/~izaguirr/papers/scicade03_2.ppt
>
> Cheers,
> Thierry
> -------------------------+------------------------------------------
>   Dr.Sc. Thierry Matthey  | work +47 55584167  | ma...@bc...
>   Para//ab, BCCS          | fax  +47 55584295  | www.matthey.org
>   Thorm=F8hlensgt. 55       | home +47 55313146  | ICQ# 23561635
>   N-5008 Bergen Norway    | www.parallab.no    | E/G/F/N
> -------------------------+------------------------------------------

Paul R Brenner, P.E.
Graduate Research Assistant
Computer Science and Engineering
The University of Notre Dame

[Protomol-users] DihedralHMC and MDLeapfrog

[Paul R B. <pbr...@nd...>]

Hello Marcelo,

One point of clarification.  I am the author responsible for the Dihedral=
HMC
integrator/propagotor code (It looks like Thierry was think of the Shadow
Hybrid Monte Carlo code when he forwarded you the ppt presentation and Sc=
ott
Hampton's contact info).

I've attached a brief description of the integrator.  An example configur=
ation
file and necessary inputs can be found with the source code in the exampl=
es
directory:

protomol/examples/butane_4/UA_butane.dhmc.conf

The new user's guide for 2.0.3 is in editing.  The users guide for 1.8.3 =
has
most of the the information you need and can be found in the source code
directory:

protomol/doc/UserGuide/

Regards,
Paul

Paul R Brenner, P.E.
Graduate Research Assistant
Computer Science and Engineering
The University of Notre Dame

Quoting Thierry Matthey <Thi...@bc...>:

> Dear Marcelo,
>
> > I would like to have access to the some documentation on Protomol ver=
sion
> 2.0.3, mainly
> > with relation to the DihedralHMC and MDLeapfrog subroutines. Where I =
can
> find them?
> > One more time thanks for the attention,
>
> I forward your request to the developers of the methods:
>
> MDLeapfrog (Ma Qun):
> http://www.cs.njit.edu/~qma/pub/42356.pdf
>
> DihedralHMC (Hampton Scott):
> http://www.nd.edu/~izaguirr/papers/scicade03_2.ppt
>
> Cheers,
> Thierry
> -------------------------+------------------------------------------
>   Dr.Sc. Thierry Matthey  | work +47 55584167  | ma...@bc...
>   Para//ab, BCCS          | fax  +47 55584295  | www.matthey.org
>   Thorm=F8hlensgt. 55       | home +47 55313146  | ICQ# 23561635
>   N-5008 Bergen Norway    | www.parallab.no    | E/G/F/N
> -------------------------+------------------------------------------

[Protomol-users] Re: Windows Executable

[Paul R B. <pbr...@nd...>]

Steve,

I typically do everything on Linux and apologize that the framework does not
currently have an embedded windows GUI.  The development team tried to
integrate a multiplatform GUI using the FOX toolkit a couple of years back but
this complicated the source code for our users.  One of Protomol's primary
goals is to provide a user friendly source code which can be extended by
researchers developing molecular simulation algorithms.  We therefore try to
keep the source as clean and readable as possible while continually expanding
the simulation capability.

Thierry introduced the perl/tk wrapper so that the Protomol source need not be
modified but the functionality of a windowing environment could be achieved.

Regards,
Paul

Quoting "Steven L. Guberman" <sl...@sc...>:

> Dear Paul,
>
> Thanks for your email. The functionality that you describe does indeed work
> for me. I was only hoping that the program would run in the same manner as
> other
> windows XP programs. Thanks for your help.
>
> Steve

Re: [Protomol-users] Protomol Windows Executable

[Thierry M. <Thi...@bc...>]

Hi Steve,

The Windows version on sourceforge.net was compiled on Windows 2000 using 
Visual Studio C++ 7. The executable itself is only command line based, 
which means you have to pass all arguements on a MD-DOS prompt or you can 
use TkProtoMol.pl. In order to use TkProtoMol.pl you need to install perl, 
I recommend to install perl from ActiveState. Take a look at 
http://protomol.sourceforge.net/user.html

Best Regards,
Thierry

[Protomol-users] Protomol Windows Executable

[Paul R B. <pbr...@nd...>]

Hello Steve,

I downloaded the windows executable on an Intel Pentium 4 system running Windows
XP Professional and the executable seems to run properly.  The executable should
be run from the command prompt environment.  If you run it without any input
configuration files the primary header with scroll up the screen and the
command line options will be displayed.

Is this the functionality that is not working for you or are you attempting the
graphical version as shown in the screen shots on the home page?  The screen
shots which show operation in a GUI window are enabled by a perl/Tk GUI
wrapper.  If the command line version works fine on your machine and this is
the functionality you are looking for let me know and either Thierry or I will
provide more details.

Regards,

Paul R Brenner, P.E.
Graduate Research Assistant
Computer Science and Engineering
The University of Notre Dame

[Protomol-users] ProtoMol v2.0 Release

[Paul R B. <pbr...@nd...>]

We are pleased to announce the release of ProtoMol v2.0: a free, open source,
object-oriented, component-based framework for molecular dynamics (MD)
simulations.  It supports CHARMM 19 and 28a2 force fields and is able to read
and write PDB, PSF, XYZ, DCD, and several common graphics formats.  ProtoMol is
designed for high flexibility, ease of extendibility and maintenance, and high
performance and scalability demands.  We use MPI to implement atom- and
force-based decompositions in a manner nearly transparent to both end-users and
developers.  The use of fast electrostatic force evaluation algorithms like
particle mesh Ewald (PME), an O(n log n) method, and multigrid summation (MG),
an O(n) method, further improve performance.  In addition, MG has superior
scalability when running in a parallel environment.

ProtoMol contains many novel MD and Monte Carlo (MC) based methods suitable for
simulation in a variety of ensembles.  For example, shadow hybrid Monte Carlo
(SHMC) has been shown to scale with system size and MD timestep by a factor of
10 or more over hybrid Monte Carlo (HMC).  Multiscale simulations take advantage
of longer timesteps by coupling multiple time-stepping algorithms with advanced
integrators.  Using targeted Molly, we have achieved stable timesteps of 16fs
for solvated proteins.  In addition, ProtoMol has been designed to interact with
VMD, a visualization engine developed at the University of Illinois, which is
used for displaying large biomolecular systems and running simulations in 3-D.

Highlights of ProtoMol v2.0:
- Average 50% performance increase versus ProtoMol 1.8.3
- Novel integrators including Shadow Hyrid Monte Carlo (SHMC),
  isothermal-isobaric MD, and isomolar-semigrand MD.
- An updated interface for adding new forces, integrators, and output methods
- Improved code organization utilizing design patterns

We invite you to visit the ProtoMol home page, http://protomol.sourceforge.net,
where C++ source code, documentation, and prebuilt binaries may be found.
Operating systems natively supported include Linux, SunOS, Irix, HP-UX and
Windows.

We have also configured a new mailing list to better serve our user base.
Please visit the following link and join the list for support and development
news:

http://lists.sourceforge.net/lists/listinfo/protomol-users

Protomol user and development questions can be sent to the list using the
following email address:

pro...@li...

For questions or comments, please feel free to contact the ProtoMol team at
pro...@cs....