Date: Tue, 26 Nov 2002 15:51:31 -0800
From: "David A. Case"
Subject: Re: solvateOct

On Tue, Nov 26, 2002, Kristina Furse wrote:

> If anyone else tries to do this, I thought I'd mention for the record, I
> ended up going with the first, more general protocol (details below)--two
> passes through LEaP, then copied the box information from the first prmtop
> file to the second (%FLAG BOX_DIMENSIONS section), and then used this second
> modified prmtop with the first coordat (since both were 3pt water models and
> the first one had the correct box info) for the calculation.

This sounds correct. Amber 7 ignores the BOX_DIMENSIONS in the prmtop (it's
just there for backwards compatibility), but it doesn't hurt.

>
> I initially tried the second, simpler way, but couldn't get it to work. I
> even used an edited leaprc.ff99 where I went in and replaced any mention of
> T3P with SPC. From this I would get SPC for the crystallographic water I had
> included, but still got T3P for the additional box water. I took a look at
> solvents.lib, and please correct me if I'm interpreting it wrong, but I
> think WATBOX216 may be specific to TIP3P after all. The entry for WATBOX216
> in solvents.lib seems to have the TIP3P charges included in the
> "!entry.SPC.unit.atoms table str name str type int typex int resx int
> flags int seq int elmnt dbl chg" section,

..oops....you are correct, and my earlier suggestion was wrong. As you
note, WATBOX216 has the TIP3P charges, and hence *cannot* be used for
SPC/E water. Thanks for catching this. (And, of course, WATBOX216 is an
archaic name, back to the days when the only water model really supported
by amber was TIP3P. We'll probably have to make WATBOX216 go away in the next
release, to be replaced with a better name.)

[And, a good object lesson for everyone: just because my initials are the
bottom of an e-mail doesn't make the information correct. Please follow
Kristina's example and check things for yourself!]


Thanks again for the information....regards....dac