Hi Anna and all others,
I could get rid of this problem by changing few parameters..
I read in amber tutorials that a time step of half femtosecond can improve
the initial stability during equilibration..
reference :
" The half femtosecond time step is probably overkill but it will improve
the chances of our simulation remaining stable early on." (
http://ambermd.org/tutorials/advanced/tutorial5_amber11/section4.htm)
I changed this parameter and it resolved the problem of vlimit exceeding at
least for me..
the new input file i used for heating up the system is
Initial heating and equilabration
&cntrl
imin=0, irest=0,
nstlim=100000, dt=0.0005,ig=-1,
ntpr=100, ntwx=10,
ntc=1, ntf=1,
ntr=1, restraintmask=':1-356',
restraint_wt=1.0,
cut=16.0, ntb=1,
ntt=3, gamma_ln=1.0,
tempi=0.0, temp0=300.0,
/
On Thu, Oct 11, 2012 at 2:13 PM, Anna Bauß <
anna.bauss.physchem.uni-freiburg.de> wrote:
> Hey,
>
> could you keep me informed about your conclusions on these problems you
> have? I have to cope with the exact same problem and would highly
> appreciate if I could get any impact on that.
>
> Thanks a lot!
>
> AnnA
>
> Am 10.10.2012 14:24, schrieb Jitesh Doshi:
>
> Hi David,
>> Thanks a lot.
>>
>> Yes.. The problem is there..the starting energy in step 0 of md is much
>> higher .. and nowhere near the energy in last step of minimization. How
>> should I look into the matter?
>> Yes i do have one non-standard residue.. my ligand..
>> Also when i visualized the structures after minimization, and what i
>> observed is the liagnd molecule loses its original connectivity. And the
>> new ligand in minimized structure is distorted and broken. Why this could
>> be happening?
>> I'll check the smaller MD run and visualize the trajectory.
>> On Wed, Oct 10, 2012 at 4:46 PM, David A Case <case.biomaps.rutgers.edu>*
>> *wrote:
>>
>> On Wed, Oct 10, 2012, Jitesh Doshi wrote:
>>>
>>> I am running minimization in three steps, with PBC on,
>>>> 1. all atoms restrained
>>>> 2. only protein backbone atoms restrained
>>>> 3. all atoms free
>>>>
>>>> all with 2500 steepest descent and 5000 conjugate gradient steps.
>>>>
>>>> and then moving to run heating and equilibration with following input
>>>>
>>> Make sure the energy at step 0 of the dynamics matches the last step of
>>> minimization. Make sure that checkoverlap (in ptraj) reports no bad
>>> contacts.
>>>
>>> Run a short (serial) simulation with ntpr=1, ntwx=1, nstlim=100.
>>> Visualize
>>> your trajectory and look for suspicious behavior.
>>>
>>> You have a big system, and this may not be easy. If there is any
>>> non-standard
>>> part of your setup (not just ordinary amino acids...) look in that region
>>> first.
>>>
>>> ...dac
>>>
>>>
>>> ______________________________**_________________
>>> AMBER mailing list
>>> AMBER.ambermd.org
>>> http://lists.ambermd.org/**mailman/listinfo/amber<http://lists.ambermd.org/mailman/listinfo/amber>
>>>
>>>
>>
>>
>
> --
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
>
--
Jitesh V. Doshi
Research Associate,
Bioinformatics Centre,
University of Pune.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Oct 18 2012 - 05:00:04 PDT
