I feel like something has changed with antechamber... although I'm not immediately seeing anything in the git revisions that would change this. Maybe I'm somehow using it differently (?). I don't know. Here's what I'm seeing though.
I'm using 0541586bf4052136921191596329ba34a466c65f
configured with gnu
gcc (GCC) 4.6.3 20120306 (Red Hat 4.6.3-2)
I have a pdb file generated by openbabel (attached).
When I run
antechamber -i PM3BP.U.pdb -fi pdb -o PM3BP.U.mol2 -fo mol2 -c bcc -s 2 -nc 0
It produces PM3BP.U.mol2, but the residue name is left blank in the mol2 file.
When I run parmchk, I get a memory allocation error
parmchk -i PM3BP.U.mol2 -f mol2 -o PM3BP.U.frcmod
memory allocation error for *bond_array
When I load the mol2 file into tleap, it dies with
Loading parameters: /home/giese/devel/git/Programs/Amber11/dat/leap/parm/gaff.dat
Reading title:
AMBER General Force Field for organic molecules (Version 1.7, Nov 2013)
Loading Mol2 file: ./PM3BP.U.mol2
Reading MOLECULE named 12 12 1 0 0
Fatal Error: last line read: 1 C -1.2390 -0.4190 0.0010 c 1 0.803500
=============
The solution is to explicitly specify the -rn flag when using antechamber.
When -rn is explicitly used, parmchk no longer produces a memory allocation error and tleap runs fine.
I was surprised by this because I've never had to use that flag before.
antechamber -h says:
-rn residue name, overrides input file, default is MOL
which suggests that the residue name should default to MOL.
I'm not seeing anything weird in the pdb file that would blank it.
...but maybe there is (?)
-Tim
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Tue Jan 07 2014 - 10:30:02 PST
