4 fs time steps with HMR is still a "classical" MD simulation.
I think the HMR article discusses the impact of longer time steps. In some
systems you may be able to get away with 2.5 fs using a strong thermostat,
but I would carefully check the energy profile and other metrics to look
for instabilities. I would be a little surprised if the energy profile
matched that from a smaller time step, but I suppose some systems are less
sensitive than others.
On Tue, Apr 20, 2021 at 2:13 AM <h_sahakyan.mb.sci.am> wrote:
> Dear Dr. Simmerling
>
> Thank you for your reply,
> HMR or virtual sites definitely allow 4fs, however, my question is about
> "classical MD". I ran MD with simple h-bonds constraints and 2.5 fs time
> step, the results of the simulation, including energies of the system were
> similar. I am curious we use 2 fs if there is no difference? Or there are
> pitfalls I have not noticed?
>
> All the best,
> Harutyun
>
>
> > Date: Mon, 19 Apr 2021 09:02:59 -0400
> > From: Carlos Simmerling <carlos.simmerling.gmail.com>
> > Subject: Re: [AMBER] 2.5 fs integration time step
> > To: AMBER Mailing List <amber.ambermd.org>
> > Message-ID:
> > <CAGk3s-SLR7c-sz9d8zK0Dh39pdimt=EpfD6PWeFxgjSE3VJXkQ.mail.gmail.com>
> > Content-Type: text/plain; charset="UTF-8"
> >
> > if you use hydrogen mass repartitioning (in parmed) then you should be
> able
> > to use a 4fs time step, we do this for most of our simulations.
> > https://pubs.acs.org/doi/10.1021/ct5010406
> >
> > On Mon, Apr 19, 2021 at 8:49 AM <h_sahakyan.mb.sci.am> wrote:
> >
> >> Dear Colleagues,
> >>
> >> Some researchers use the 2.5 fs integration time step, an example
> >> is this work How Fast-Folding Proteins Fold | Science (
> >> https://science.sciencemag.org/content/334/6055/517/tab-article-info).
> Do
> >> you have any ideas on how 2.5 fs can affect MD simulations?
> >>
> >> All the best,
> >> Harutyun Sahakyan
> >> _______________________________________________
> >> 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 Tue Apr 20 2021 - 13:00:02 PDT
