Re: scnb value to prevent VDWs scaling

From: Marc Gordon (marcgrdn55_at_gmail.com)
Date: Mon Sep 10 2012 - 08:28:25 CDT

Hi Axel thanks for the prompt response.

On Mon, Sep 10, 2012 at 3:02 PM, Axel Kohlmeyer <akohlmey_at_gmail.com> wrote:

> On Mon, Sep 10, 2012 at 2:56 PM, Marc Gordon <marcgrdn55_at_gmail.com> wrote:
> > Hi Axel
> >
> > The NAMD documentation for sure mentions scnb but as I said it doesn't
> > explicitly say that an scnb value of 1 results in no scaling. Just that
> the
> > vdws interactions are divided by the scnb values.
>
> for AMBER inputs.
>
> [...]
>

Wait... for AMBER inputs?

So scnb is not even considered unless the 'amber' option is set to 'yes'?
.. so it is defaulted to 1 with charmm inputs?

>
> > On a related note then do you guys happen to have any idea as to why NAMD
> > would scale vdw interactions to half by default? (default value of scnb
> is
> > 2)
>
> because that is what amber uses most commonly.
> (1-4 LJ scaled by 0.5 and 1-4 coulomb scaled by 5/6th)
>
> axel.
>

Ah OK. I'm using GLYCAM and not AMBER (though it uses the same file
formats). I chatted to one of the maintainers and he said GLYCAM doesn't
want scaling either. I tried it on both the default (2) and set to 1.
Messing with it is what ultimately brought me to asking this question
because I thought "well if GLYCAM wants no scaling and assuming CHARMM
wants no scaling why would the default be 2?".

>
> >
> >
> > Marc
> >
> >
> >
> > On Sat, Sep 8, 2012 at 1:04 PM, Axel Kohlmeyer <akohlmey_at_gmail.com>
> wrote:
> >>
> >> On Sat, Sep 8, 2012 at 12:21 PM, Marc Gordon <marcgrdn55_at_gmail.com>
> wrote:
> >> > HI Norman
> >> >
> >> > Yes that is what I thought but I couldn't find it explicitly confirmed
> >> > in
> >>
> >> actually it is *not* quite that.
> >>
> >> > the NAMD documentation and didn't want to just assume I had it right--e89a8f234589ac0bcd04c958ee82--

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:22:03 CST