I would say that this depends very much on what you are trying to learn,
and what information you have to start.
Different properties converge at different rates, so the length of the MD
should ideally be several times longer than the timescale of the properties
you are investigating. Also, if any part of your initial structure is a
model, you will want/need to show that the results do not depend on how you
built that model. For example, if you added part of the structure
(glycans?) you'll want to add them in a different conformation and make
sure that any conclusions drawn from the simulation are not strongly
dependent on how you made your initial model. Simply repeating the
simulation from the same model will not explore that sensitivity, since it
will just tell you that the results are the same if you start with the same
model.
hope that helps .. there is not really 1 simple answer.
On Fri, Oct 20, 2017 at 7:28 AM, Lizelle Lubbe <LBBLIZ002.myuct.ac.za>
wrote:
> Hi All,
>
> I am aiming towards 30ns of production dynamics and have been advised to
> repeat this 3 times to demonstrate data reproducibility but am not sure how.
> As suggested in the manual, I've used ig=-1 throughout heating,
> equilibration and production - is this correct?
>
> Here's the input file for the first section of production, restarting from
> equilibration.rst7 and reading its velocities.
>
> First Dynamic Simulation with Constant Pressure
>
> # Control section
> &cntrl
> imin = 0,
> irest = 1,
> ntx = 5,
> ntb = 2, ntp = 1, pres0 = 1.0, taup = 1.0,
> cut = 10.0,
> ntc = 2, ntf = 2,
> temp0 = 300.0,
> ntt = 3, gamma_ln = 2, ig = -1,
> nstlim = 500000, dt = 0.002,
> ntwx = 5000, ntpr = 5000, ntwe = 5000, ntwr = -100000,
> ioutfm = 1,
> iwrap = 1,
> ntr = 0,
> ntwprt = 0,
> nmropt = 1,
> /
> &wt type = 'END'/
> DISANG = NDOM_rnb.RST
> /
>
> There's a wall-clock time-limit on our HPC so the simulations were
> restarted each time using the input given above until the desired
> trajectory length was reached (always irest=1, ntx=5 and ig=-1).
>
> I don't really understand the concept of MD reproducibility very well...
> Could someone please explain this to me?
>
> Currently, I can think of two ways for restarting the production but don't
> know which is correct:
> 1. using ig=-1, irest=0, ntx=1 and starting from the same initial
> equilibrated coordinates as used for the first production
> 2. using ig=-1, irest=0, ntx=1 but starting from the rst7 file saved at
> 10ns for the second and 20ns for the third repeat production
>
> Would option two demonstrate data reproducibility or just explore more
> conformational space (or are they the same thing)?
> Is something like 30ns generally acceptable thesedays to investigate the
> motion of glycans and protein breathing or should it be increased?
>
> Kind regards
>
> Lizelle Lubbe
>
> PhD (Medical biochemistry) candidate
> Department of Integrative Biomedical Sciences
> University of Cape Town
> Disclaimer - University of Cape Town This e-mail is subject to UCT
> policies and e-mail disclaimer published on our website at
> http://www.uct.ac.za/about/policies/emaildisclaimer/ or obtainable from
> +27 21 650 9111. If this e-mail is not related to the business of UCT, it
> is sent by the sender in an individual capacity. Please report security
> incidents or abuse via csirt.uct.ac.za
>
> _______________________________________________
> 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 Fri Oct 20 2017 - 05:30:03 PDT
