I agree with Bhavaraju:
It is unlikely that the difference in the RMSF comes from use if
iwrap. As I mentioned in my first reply, your heating is indeed too
fast, so it is very likely that none of your calculations is
converged.
First, 20 or even 40ps is too small a time for heating. Myself, I
usually do that in at least 2 steps: 100ps to take the system from
0-100K, then another 100-150ps to take the system from 100 to 300K
(NVT). Of course, that can vary depending on the system.
Then comes the constant pressure part. You don't mention how long that
equilibration was, but it should be long, probably longer than the NVT
simulations. For some systems, for example, I have used up to 2ns, but
that again can depend on the system.
Finally, you also don't mention what are the conditions of the final
15ns simulations. Is it NPT, NVT, NVE, etc?
Cheers,
Gustavo Seabra.
On Mon, Aug 23, 2010 at 5:03 PM, M. Reza Ganjalikhany
<ganjalikhany.gmail.com> wrote:
> Thank you for your email.
>
> I'm looking for local fluctuations in a psychrophilic enzyme at different
> temperatures (278,308 and 333). I found a meaningful difference at the first
> experiment which was not reproduced on the iteration.
>
>
>
> #1: (iwrap=0)
>
> A- Heating system with weak restraints on protein in 20 ps with constant
> volume from 0 K to (278,308 and 333) K
>
> B- Equilibration at constant pressure at (278,308 and 333) K
>
> C- Three sets of 15 ns MD simulation at (278,308 and 333)K
>
>
>
> #2: (iwrap=1)
>
> A- Heating system with weak restraints on protein in 40ps (constant volume)
> from 0 K to (278,308 and 333) K
>
> B- Equilibration at constant pressure at (278,308 and 333) K
>
> C- Three sets of 15 ns MD simulation at (278,308 and 333)K
>
>
> Finally I found RMSFs from experiment #1 completely different with #2 at
> certain positions. On the other words, the fluctuation at certain positions
> has been decreased with the temperature in #1 which was reversed in #2.
>
> The differences in these experiments are including the heating time and the
> imaging (iwrap).
>
>
>
> RMSFs are attached.
>
>
>
> Regards,
>
> M.Reza
>
>
>
> On Mon, Aug 23, 2010 at 7:26 AM, Gustavo Seabra <gustavo.seabra.gmail.com>wrote:
>
>> Although it is true that you may be using a equilibration time that is
>> too small, and that your heating is too fast, especially since you are
>> using explicit solvent, the main difficulty in answering your question
>> is that you also don't mention what properties you are looking at. How
>> did you conclude that your results after 15ns are "completely
>> different"? What properties did you measure? How do they differ?
>>
>> Note that, even with the exactly the same protocol, the precise
>> trajectories and final structures are expected to differ between two
>> different simulations.
>>
>> Cheers,
>> Gustavo Seabra
>> Professor Adjunto
>> Departamento de Química Fundamental
>> Universidade Federal de Pernambuco
>> Fone: +55-81-2126-7417
>>
>>
>>
>> On Mon, Aug 23, 2010 at 6:01 AM, M. Reza Ganjalikhany
>> <ganjalikhany.gmail.com> wrote:
>> > Dear all,
>> >
>> > I have done two similar MD simulations with a 300 a.a. protein in
>> explicit
>> > solvent with different equilibration time.
>> > The first equilibration time was 20ps and the second was 40 ps (from 0 K
>> to
>> > 300 K) and then I got completely two different results after 15ns MD
>> > simulations.
>> >
>> > Is it normal to obtain such results? How long should be sufficient for
>> such
>> > system to be equilibrated?
>> >
>> > Any help would be greatly appreciated.
>> >
>> > Regards,
>> > M. Reza
>> > _______________________________________________
>> > AMBER mailing list
>> > AMBER.ambermd.org
>> > http://lists.ambermd.org/mailman/listinfo/amber
>> >
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Aug 23 2010 - 18:30:08 PDT
