Re: [AMBER-Developers] sleap -> tleap migration

From: B. Lachele Foley <>
Date: Fri, 18 May 2012 01:08:38 +0000

Sorry I haven't weighed in before now. I haven't read all bits of the conversation either. Apologies if I duplicate. (many hardware failures lately)

The only thing to be wary of with using sleap is that it might still be a bit buggy. We like what sleap was trying to do. We just kept having trouble building files with sleap and gave up. Granted, there are issues in tleap, but those are long known, and we had already learned to deal with them. We finally made our own builder. It makes topology files. and are the development sites. The code there is still a bit new, but being used and tested regularly at

:-) Lachele

Dr. B. Lachele Foley
Complex Carbohydrate Research Center
The University of Georgia
Athens, GA USA

From: Eric Pettersen []
Sent: Thursday, May 17, 2012 4:33 PM
To: AMBER Developers Mailing List
Cc: Wei Zhang; zhangqian
Subject: Re: [AMBER-Developers] sleap -> tleap migration

On May 17, 2012, at 6:06 AM, David A Case wrote:

> On Mon, May 14, 2012, Eric Pettersen wrote:
>> Are you suggesting sticking with AmberTools11, or instead using
>> AmberTools12 but copy sleap into it? If the former, then obviously I
>> wouldn't benefit from any improvements to antechamber/sqm [which
>> Chimera also uses] and wouldn't be able to offer the new ff12SB force
>> field. If the latter (assuming that would work), there'd still be
>> the
>> drawback that generated prmtop files wouldn't have any of the new
>> sections that tleap writes, but maybe that's not a big deal -- I
>> don't
>> know....
>> You are right about tleap and mol2. I did some of the work on the
>> conversion to Amber12 before running into the parmchk thing, and for
>> Solvate I had to change to providing PDB as input for tleap since
>> tleap had no idea what the atom types were if I provided mol2.
> Can you elaborate? This doesn't (seem to) accord with my
> experience: the
> loadMol2 function in tleap pulls in atom types correctly, at least
> for me.
> But maybe the mol2 files I'm using are different from the ones you
> have(?)

Hi Dave,
        I'm guessing your mol2 files are from antechamber, are for small
molecules, and contain GAFF atom types. For Solvate, obviously the
systems are typically proteins or nucleic acids. For those, if tleap
is provided a standard mol2 file with Sybyl atom types, e.g.:

       1 N 49.6680 24.2480 10.4360 N.4 1
HIS1 0.0000
       2 CA 50.1970 25.5780 10.7840 C.3 1
HIS1 0.0000
       3 C 49.1690 26.7010 10.9170 C.2 1
HIS1 0.0000

then it will not know anything about the atom types and for each and
every heavy atom will emit a message like this during the solvation

(using default radius 1.500000 for CB)

followed later by this (single) message:

(type - hence mass - of one or more atoms could not be found)

This stands in contrast to tleap's handling of PDB files of proteins
and nucleic acids where it figures out the atom types. I guess the
philosophy is "big systems via PDB files; small molecules via mol2",
which is fine. Anyway, that's why I had to switch around to using PDB
as input to tleap in the Solvate plugin -- I guess sleap does some
magic for mol2 files so that they work fine for sleap for this purpose
> Generally, I'd think that copying sleap into AmberTools12 ought to
> mostly
> work. Of course, it would be great if good features of sleap could
> be ported
> to tleap; (or vice-versa, but [as we have seen], keeping two
> versions of the
> LEaP functionality in sync requires a real development commitment.).

This seems like the way to go. It involves the smallest change to the
plugins, and at least allows me to offer the latest force fields. The
WritePrmtop plugin will write out an older form of the prmtop file,
but perhaps if Qian takes over sleap development then that can be


AMBER-Developers mailing list

AMBER-Developers mailing list
Received on Thu May 17 2012 - 18:30:03 PDT
Custom Search