Well-tempered ensemble in PLUMED2 when combined with parallel tempering

[juve...@gmail.com]

Dear all,

I tried to perform parallel tempering metadynamics on my system using PLUMED2. I also tried to implement well-tempered ensemble by using potential energy as a CV. However the output of parallel tempering in GROMACS 4.5.5 only gives one COLVAR and one HILL files, which are for the ensemble at the lowest temperature. 

I checked the document for METAD action in PLUMED2. It states that if using BIASFACTOR one has to explicit specify temperature. However when passing PLUMED input to GROMACS I can only use one plumed.dat file, which means one temperature. Does that mean the MD engine only do metadynamics on the lowest temperature ensemble? Could that be the reason why I am getting only one COLVAR, instead of N files with N being number of replicas? I am aware that PLUMED1.3 has no problem in providing N COLVAR files for each temperature replica.

My plumed.dat is as follows:

ENERGY LABEL=ene 

METAD ARG=ene SIGMA=1000 HEIGHT=1 PACE=500 BIASFACTOR=10 TEMP=300 LABEL=restraint

PRINT ARG=ene,restraint.bias STRIDE=500 FILE=COLVAR

And my mdrun command is:

$HOME/compiling/gromacs-4.5.5/install_dir_plumed/bin/mdrun_mpid -v -s topol_.tpr -multi 19 -replex 1000 -reseed -1 -plumed plumed.dat

Your kind help is greatly appreciated!

Weilong

[Massimiliano Bonomi]

Dear Weilong,
metadynamics is carried out at all temperatures, which are passed to PLUMED by GROMACS.
I think you should execute the mdrun command using the "mpirun" shell script (or similar ones, depending
on the machine you are using). Typically it should something like this:

mpirun -np 19 $HOME/compiling/gromacs-4.5.5/install_dir_plumed/bin/mdrun_mpid -v -s topol_.tpr -multi 19 -replex 1000 -reseed -1 -plumed plumed.dat

Max

> --
> You received this message because you are subscribed to the Google Groups "PLUMED users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
> To post to this group, send email to plumed...@googlegroups.com.
> Visit this group at http://groups.google.com/group/plumed-users.
> For more options, visit https://groups.google.com/d/optout.

[Giovanni Bussi]

Hi,

notice that in plumed 2 the temperature is *not* passed by gromacs. You should prepare several different plumed.dat files, named e.g.

plumed.dat.0

plumed.dat.1

etc

Giovanni

[juve...@gmail.com]

Thank you for the kind reply! I am sorry I did not put mpiexec in my post but it was actually there. I guess the problem was GROMACS did not pass temperature for other replicas into PLUMED2, just as Giovanni mentioned. 

Thanks in advance.

Weilong 

[juve...@gmail.com]

Right. If I am understanding correctly METAD action only use the explicitly specified temperature. I know PLUMED1.3 can do parallel tempering without multiple plumed.dat. Is there any reason to change that in PLUMED2?

Thanks,

Weilong

[Giovanni Bussi]

PLUMED2 searches first for file plumed.dat.X, where X is the number of the replica. In case it is not found, it searches for plumed.dat

Thus, I think that with multidir you will not need the different suffixes (but I never tried in practice).

Giovanni

On Fri, Apr 18, 2014 at 12:38 AM, <sebastie...@gmail.com> wrote:

Is this needed even if we use the -multidir option of mdrun?

Thanks,

Sebastien

On Thursday, April 17, 2014 9:10:38 AM UTC-4, Giovanni Bussi wrote:

[Sathish Dasari]

Hi,

I was trying Parallel Tampering Metadynamics in Well Tempered Ensemble (PTMetaD-WTE) tutorial. 

1) As there are four replicas and four HILLS files (HILLS_PTMetaDWTE.0, HILLS_PTMetaDWTE.1, HILLS_PTMetaDWTE.2, HILLS_PTMetaDWTE.3), which HILLS file do we use to reconstruct the free energy profile.

2) Once we get the free energy profile, from where to get the corresponding conformations of minima? As there are four trajectories corresponding to WTE and four trajectories from PTMetad on WTE.

Thank you