[gmx-users] question about Polak Ribiere conjugate gradient

[Erik Lindahl]

Hi,

1e-6 kJ/mol is a very tough energy tolerance for a big system :-)

In general, if you apprioximate the energy surface close to the minimum  
with harmonic potentials, you cannot do better than sqrt(machine  
precision) in both the coordinate and value.

Considering a protein consists of thousands of degrees of freedom  
you'll have some rounding errors too, so I would be very surprised if  
you could get to 1e-6.

And, remember, all you can hope for is a *local* minimum. With that in  
mind there is rarely any reason to go lower than 0.1 kJ/mol...

Cheers,

Erik



On May 3, 2004, at 4:49 PM, David wrote:

> On Mon, 2004-05-03 at 14:08, Hua Wong wrote:
>> I am doing a minimization using the brand new GROMACS 3.2. It works  
>> fine on a system that previously lead to error with GROMACS 3.1.
>>
>> I am working on a protein minimised under NAMD and wanted to minimise  
>> it further using GROMACS to correct some omega torsion problems. All  
>> seems to be fine, only I obtain a warning :
>>
>> Stepsize too small, or no change in energy.
>> Converged to machine precision,
>> but not to the requested precision Fmax < 1e-06
>>
>> I am using CG with GROMACS compiled in double precision.
>>
> try using
> integrator = l-bfgs
>
>
>> Here is the .mdp file I use.
>> ##################################
>> title                    = GMXminiomega
>> cpp                      = /lib/cpp
>> include                  = -I../top
>> define                   = -DFLEXIBLE
>> integrator               = cg
>> emtol                    = 0.000001
>> emstep                   = 0.01
>> nstcgsteep               = 100
>> nsteps                   = 3000
>> nstxout                  = 50
>> xtc_grps                 = Protein
>> energygrps               = Protein
>> nstlist                  = 5
>> ns_type                  = grid
>> rlist                    = 1.
>> vdwtype                  = cut-off
>> coulombtype              = cut-off
>> rcoulomb                 = 1.
>> rvdw                     = 1.
>>
>> Is there anything I've forgotten ?
>>
>> Thanks
>> _______________________________________________
>> 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.
> -- David.
> _______________________________________________________________________ 
> _
> David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
> Dept. of Cell and Molecular Biology, Uppsala University.
> 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.