i agree that you should be using a series of simulations with slow heating
and gradual release of restraints. if not, it's not at all surprising that
some simulations completely unfold.
2010/4/13 Jianing <song_jianing_hi.126.com>
> Hi Bill,
> After the system has been heated to 300K with the solute fixed in 100ps,
> the keyword 'irest=1' instead of 'irest=0' should be reasonable when perform
> the next equlibration for 5ns. I have checked the structure after
> equlibration, it sounds reasonable, similar to the XRD structure of villin
> headpiece. But I will change the value of irest to 1 to see the difference
> and maybe this is the reason for my strange results, you know,
> small error accumulates into big mistake.
> >Normally one equilibrates by slow warming and/or keeping the
> >solute fixed with restraints.
> what's the exact meaning of this? Do you mean that equlibration should
> performed with restraints generally? But equlibration is to make the system
> equilibrated and don't corelate to the initial structure. If you put
> constraints on the system, is this too manual and don't leave the system
> where it is.
> Thank you in advance.
> Jianing
>
>
>
>
>
>
>
> 在2010-04-14 00:57:28，"Bill Ross" <ross.cgl.ucsf.edu> 写道：
> >This mdin seems like a fast-warming regime that I would hesitate to
> >even call equilibration:
> >
> > irest = 0,
> > ...
> > tempi = 300.0,
> > temp0 = 300.0,
> >
> >Have you looked at your structure to see if it's reasonable?
> >Normally one equilibrates by slow warming and/or keeping the
> >solute fixed with restraints.
> >
> >It will be interesting to see how replicable your results are
> >in the parallel execution situation.
> >
> >Bill
> >
> >_______________________________________________
> >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 13 2010 - 20:00:03 PDT
