Re: Rigid bonds and hydrogen bonds "on" inaccurate/invalid?

From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Wed Mar 19 2008 - 15:41:14 CDT

As far as I can see (and people can feel free to correct me if this
seems wrong), running with langevinhydrogen on and rigidbonds on will
not affect *thermodynamic* properties, but it will affect the dynamics
-- from what I saw in my tests, you're likely to see slower dynamics,
and this will certainly affect observed rates, and could also make
sampling harder. So as long as you're comfortable with how much sampling
was done (keeping in mind the longer relaxation times) I don't think
there will be anything qualitatively wrong with the conformations you've
observed, but I wouldn't even try to make statements on rates (not that
you really should anyway, if using langevin dynamics), and ideally one
should do everything with langevinhydrogen off if you're using rigidbonds.

Best,
Peter

Kamilla Kopec wrote:
> Dear Peter,
>
> Thanks for the feedback.
>
> Ultimately, I am trying to look for conformational changes in my
> covalently bound substrate over the course of the trajectory. So I
> guess my question really is - would the combination of "langevin
> hydrogen on" and rigid bonds affect conformational preferences? Could
> a trajectory generated in this way give a qualitative idea of the
> stable conformers adopted by a substrate?
>
> A number of my colleagues have several trajectories generated in this
> way. Where possible we are aiming to repeat our work with optimal
> settings. However, where this is not possible we are trying to work
> out what caveats should be placed on the interpretation of our
> trajectories. As above, our main interest is substrate conformational
> change.
>
> Any comment would be much appreciated.
>
> Best Regards
>
> Kamilla
>
>
> */Peter Freddolino <petefred_at_ks.uiuc.edu>/* wrote:
>
> Well, depends on what you mean by "accuracy". If you have
> langevinHydrogen on with rigidbonds you will have calculated
> *something*, but many physical properties will be different. I tried
> this experiment once with water; I believe I got a diffusion constant
> that was about 4x smaller using the same gamma if langevinHydrogen
> and
> rigidbonds were both on, and some other values came out different as
> well, so I wouldn't recommend mixing these trajectories with those
> that
> have been run properly.
> Hope this helps...
> Best,
> Peter
>
> Kamilla Kopec wrote:
> > *I mean Langevin "hydrogen off" not "hydrogen bonds off".
> >
> > */Kamilla Kopec /* wrote:
> >
> > Hi,
> >
> > I didn't pick up on the fact that "hydrogen bonds off" should be
> > used with "rigid bonds all" during my set up and I have a couple
> > of trajectories generated with the setting rigid bonds all and
> > hydrogen bonds on. Could someone clarify the implications of this
> > for the accuracy of the trajectories I have.
> >
> > Best Regards
> >
> > Kamilla
> >
> >
> >
> ------------------------------------------------------------------------
> > Rise to the challenge for Sport Relief with Yahoo! for Good
> >
> >
> >
> >
> ------------------------------------------------------------------------
> > Rise to the challenge for Sport Relief with Yahoo! for Good
> >
>
>
> ------------------------------------------------------------------------
> Rise to the challenge for Sport Relief with Yahoo! for Good
> <http://us.rd.yahoo.com/mailuk/taglines/isp/control/*http://us.rd.yahoo.com/evt=51947/*http://uk.promotions.yahoo.com/forgood/>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:49:20 CST