Hi all,
I've been trying to run a mol2 file containing a methylimidazole (a histidine fragment) through antechamber and ascribe GAFF atom types and parameters to it.
Unfortunately, when I try, I get this message:
Running: /Applications/Amber11/2010-04-18/amber11/bin/atomtype -i ANTECHAMBER_AC.AC0 -o ANTECHAMBER_AC.AC -p gaff
WARNING: atom type of N (nd) and C1 (cc) may be wrong
Total number of electrons: 44; net charge: 0
This happens using Antechamber with AmberTools 1.2 or 1.4.
The GAFF atom types, as listed at
http://ambermd.org/antechamber/gaff.html#atomtype, suggest that nd is indeed wrong for a nitrogen atom in a five-membered aromatic ring (nd is for a conjugated chain). Furthermore, cc is for an "inner" sp2 carbon in a conjugated ring, while the aromatic type is ca.
The article "J. Phys. Chem. A 2009, 113, 4507–4517", by Chen et al., describes atom types for imidazole, and suggests that one nitrogen atom should be na and the other nc (though they use upper-case letters in what I presume was an editorial decision). I'll guess that they're correct unless someone indicates otherwise.
The issue is not one of ring perception per se: ATOMTYPE.INF contains this section:
-------------------------------ring property (II)------------------------------------
atom[ 2] (C1 ) belongs to one member of 5-membered ring (No 1)
atom[ 3] (N ) belongs to one member of 5-membered ring (No 1)
atom[ 4] (C2 ) belongs to one member of 5-membered ring (No 1)
atom[ 5] (N1 ) belongs to one member of 5-membered ring (No 1)
atom[ 6] (C3 ) belongs to one member of 5-membered ring (No 1)
I've attached the mol2 file that I used as input; the antechamber command is as follows:
$AMBERHOME/exe/antechamber -fi mol2 -i ile44-his68-big-foldedf2.mol2 -fo prepi -o ile44-his68-big-foldedf2_ac.prepi -c bcc -s 2 -nc 0
Any help would be much appreciated. Thanks!
Ben Roberts
PS. I can, of course, manually edit the prepi (or mol2, or other antechamber output) file and re-run parmchk. But that seems a bit dicy, especially if I've guessed the wrong atom types. So I welcome other suggestions.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Apr 23 2010 - 15:00:03 PDT
