AMBER Archive (2007)

Subject: Re: AMBER: Implicit solvent simulation is too slow.

From: SeungPyo Hong (sp1020_at_gmail.com)
Date: Fri Aug 10 2007 - 09:04:28 CDT

It's my mistake.
Not the same simulation, the simulation with same protein with explicit
solvent.

I used sander.
I haven't been using pmemd yet. I will try to use it...

I just read archieves and also find your answer to aother amber user.
Also my system is so big that make the implicit solvent simulation be slower
than explicit solvent simulation.

Thank you very much!

(Anyhow, is there any comparison study about the speed of implicit and
explicit solvent simulation according to the size of protein? If there is,
let me know where to find it.)

On 8/10/07, Carlos Simmerling <carlos.simmerling_at_gmail.com> wrote:
>
> the same simulation meaning also in implicit solvent?
> Amber is not optimized for implicit solvent simulations with cutoff.
> it has been discussed many times on this list that PME simulations for
> large systems are often faster than GB simulations. Check the archives.
>
> we might be able to help more if you tell us which program you
> are using (sander, pmemd, other)
>
>
> On 8/10/07, SeungPyo Hong <sp1020_at_gmail.com> wrote:
> > Dear amber users,
> > Go to the point, I found my Implicit solvent calculation is too slow.
> > Could anybody tell me whether my simulation is problematic or Amber
> itself
> > is slower than I had expected?
> >
> > It tooks about 28.3 hours for 1ns simulation under following condition:
> >
> > 28.3 hours / 1ns simulation
> > 16 Xeon DP 2.8 GHz CPUs
> > 4107 atoms (288 residues protein + 1 ligand)
> > implicit solvent simulation
> >
> > The run option of the MD is this:
> > 1 Equilibration Molecular Dynamics: name (Implicit Solvent Model)
> > 2 &cntrl
> > 3 nstlim = 50000, dt = 0.002,
> > 4 tempi = 300.0, temp0 = 300.0,
> > 5 ntpr = 100, ntwx = 100, ntwr = 1000,
> > 6 imin = 0, irest = 1, ntx = 7,
> > 7 ntb = 0,
> > 8 taup = 2.0,
> > 9 cut = 10,
> > 10 ntr = 0,
> > 11 ntc = 2, ntf = 2,
> > 12 igb = 1,
> > 13 ntt = 3, gamma_ln = 1.0
> > 14 /
> >
> > I didn't run explicit solvent simulation yet. (It is waiting its turn in
> the
> > queue...)
> > But the same simulation done with GROMACS tell me it tooks about 15.4hour s
> > for 1ns simulation.
> >
> > --
> > --------------------------------------------------
> > 'God used beautiful mathematics in creating the world.'
> > -Paul Dirac
> > 'But he created too many object.'
> > -Seungpyo Hong
> >
> > Seungpyo Hong
> > Master student at PBIL(Protein BioInformatics Lab.) in KAIST, Daejeon
> Korea
> > Tel. (82)-18-372-2468
> > sphong__at_kaist.ac.kr
> > sp1020_at_gmail.com
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
> 

-- 
--------------------------------------------------
'God used beautiful mathematics in creating the world.'
-Paul Dirac
'But he created too many object.'
-Seungpyo Hong

Seungpyo Hong
Master student at PBIL(Protein BioInformatics Lab.) in KAIST, Daejeon Korea
Tel. (82)-18-372-2468
sphong__at_kaist.ac.kr
sp1020_at_gmail.com

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu