Re: [AMBER-Developers] impose, ff12 and glycam tests

From: B. Lachele Foley <lfoley.uga.edu>
Date: Tue, 20 Mar 2012 20:22:25 +0000

I got your email. I was pleading for a different outcome. :-)

I was saying that since every manual, at least since Amber 8, has given impose instructions without requiring quotes, and since that has worked all this time, it would be nice if it would still work. If we change it now, it will break every script involving impose that pretty much everyone has written for all these years. Even if this is how it should have been all along, it's a change.

Of course, we can make that change, and an addition to the manual is good. I just wanted to put in one more plea not to alter behavior that's been around so long. For example, although leap occasionally inverts chirality for us in ways we don't expect, I would hesitate to "fix" that because lots of people probably rely on that behavior.

In any case, I don't really understand why that confuses leap, or why ROH confused leap. It seems that leap should be able to get that information from the context. Are there ever unit's in that location in that context? I suppose I could see what the code is doing. It seems like it should be fixable.

:-) Lachele

Dr. B. Lachele Foley
Complex Carbohydrate Research Center
The University of Georgia
Athens, GA USA
lfoley.uga.edu
http://glycam.ccrc.uga.edu

________________________________________
From: David A Case [case.biomaps.rutgers.edu]
Sent: Tuesday, March 20, 2012 3:57 PM
To: AMBER Developers Mailing List
Subject: Re: [AMBER-Developers] impose, ff12 and glycam tests

On Tue, Mar 20, 2012, B. Lachele Foley wrote:

> I vote for figuring out what about ff12SB is breaking behavior that's
> been documented and working since at least Amber 8. Can that happen?

Did my email not get through from last night? (Amber dev articles from
yesterday are not yet archived, but I thought I sent it there.) Anyway, here
it is again:

After a bit of poking around, the problem is that the "C" atom in Lachele's
examples are getting mixed up in LEaP because "C" is also a UnitID (cytosine
in RNA). LEaP expects the arguments to impose to be strings, and stops with
an error message when it finds something that is supposed to be a unit.

The correct(!) way to use the impose command is to force arguments to
be strings:

> m6 = sequence { NALA ALA ALA CALA }
> impose m6 { 3 } { { "N" "C" "CA" "O" 120.0 } }

So, you are not using the impose command correctly: you need to force all
arguments to impose to be strings, as per above. Examples in the AmberTools
manual have now been updated. As far as I can see, this is not an error
in either LEaP or in leaprc.ff12SB. When I use quotes I don't see any errors
like those Lachele is reporting.

>
> PS: The reason Glycam has an "ROH" residue rather than an "OH" residue
> is because leap kept getting the residue name "OH" confused with, I
> think, the atom type "OH." I changed to ROH and, poof, it all worked.

The "poof" is exactly the same "problem" that you reported with ff12SB. LEaP
is actually being consistent here: a bare set of letters (such as OH) is only
interpreted as a string if it doesn't appear to be anything else (like a
unit). The impose command is expecting strings, so that is what it should be
given. [That is, using quotes is not a workaround or a kludge: using quotes
for the arguments to impose is what one should have been doing all along.]

...hope this helps....dac


_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers



_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Tue Mar 20 2012 - 13:30:02 PDT
Custom Search