AMBER Archive (2007)

Subject: Re: AMBER: Temperature Drift in MD (NAB)

From: adelene_at_Stanford.EDU
Date: Fri May 18 2007 - 18:56:39 CDT


Only atoms in the bases 19-26 were allowed to move by fitting
restraints in the mme_init command.

The typical input file goes as
  mm_options("cut=25.0, nsnb=100, gamma_ln=50, temp0=300, tempi=50,
gb=1, kappa=0.19");
  mme_init(molMD, ":19-26:", "::ZZZ", dummy, NULL);
  md(3*molMD.natoms, 100000, xyz, f, v, mme);
  setmol_from_xyz(molMD, NULL, xyz);
  putpdb("test.pdb",molMD);

The other parameters are kept as their default value. (time step is 1fs)

Thanks!

Adelene

Quoting "David A. Case" <case_at_scripps.edu>:

> On Fri, May 18, 2007, adelene_at_stanford.edu wrote:
>
>> Hi, in my MD simulation (freezing 1141 atoms, moving 255) using NAB
>> and options shown below, the temperature drifts from the set 300K to
>> an average of 290.5K.
>
> What do you mean by "freezing"? (ibelly?) It is certainly possible that the
> Langevin implementation has a bias with frozen atoms: I haven't tested this
> myself.
>
>>
>> I'm not sure if this is a problem due to the coupling parameters I've
>> chosen, and would appreciate some advice on this. Should I be using
>> the Berendsen coupling, or is Langevin a better option given an
>> implicit GB run?
>>
>> MD options:
>> cut=25;
>> gamma_ln=50;
>> tempi=50;
>> gb=1;
>> kappa=0.19;
>
> Above look fine to me....How about dt, and so on? Can't you give
> the complete
> input file?
>
> ....dac
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>

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