Hi,
to understand how antechamber works is to first read the original paper on it (I just don't have the reference here at the moment but it must be cited in the manual) or best probably to read the source code. If you run antechamber with -s 2 it tells what "sub-programs" it is running (antechamber itself is mostly a driver program). You would then see that the first one to run is bondtype followed by atomtype.
It is not clear to me, however, what your argument regarding the atom types really is. If you consider the pyrimidine ring to be aromatic then surely both C3 and C4 (the two fusion points) must be the same, aromatic, atom type. You have drawn it that way yourself.
Cheers,
Hannes.
________________________________________
From: Robert Molt [rwmolt07.gmail.com]
Sent: 28 April 2015 02:19
To: AMBER Mailing List
Subject: [AMBER] Antechamber Algorithm for Atom Type
Good evening!
I have a small, planar, aromatic organic molecule (C, N, H, O atoms). I
encountered a problem in parameterizing with GAFF; 4 dihedrals were not
available. I tried to identify the four atoms in the dihedral in my
molecule, so I could just calculate the torsional PES and add the
parameters, and found some challenges. I have attached a photo of the
molecule in question.
a.) Specifically, I was informed that I lacked 2 ca-ca-cc-h4 dihedral
parameters and 2 ca-ca-cc-nd dihedral parameters by xleap. However, when
I look at the molecule, this designation does not seem to make sense.
There are 3 ca type carbons in the molecule, if I understand properly,
and none of them are adjacent to one another; there is no cc adjacent to
a h4. I /assume/, perhaps erroneously, that the "best" choice in
dihedrals to form will be the ones that do not "jump" over atoms and
thus go over grater distances.
b.) To confirm this, I examined ANTECHAMBER_AC.AC. It lists the
following table:
ATOM 1 C1 MOL 1 -1.428 -1.191 -0.001 0.000000 ca
ATOM 2 N1 MOL 1 -2.039 0.009 0.000 0.000000 nb
ATOM 3 C2 MOL 1 -1.280 1.097 -0.003 0.000000 ca
ATOM 4 C3 MOL 1 0.124 0.968 -0.003 0.000000 ca
ATOM 5 C4 MOL 1 0.596 -0.349 0.001 0.000000 ca
ATOM 6 N2 MOL 1 -0.146 -1.465 0.002 0.000000 nb
ATOM 7 C5 MOL 1 1.299 1.764 -0.006 0.000000 cc
ATOM 8 H1 MOL 1 1.405 2.837 -0.014 0.000000 h4
ATOM 9 N3 MOL 1 -1.900 2.297 -0.029 0.000000 nh
ATOM 10 H2 MOL 1 -2.896 2.310 0.100 0.000000 hn
ATOM 11 H3 MOL 1 -1.380 3.137 0.142 0.000000 hn
ATOM 12 N4 MOL 1 2.371 1.010 -0.003 0.000000 nd
ATOM 13 N5 MOL 1 1.942 -0.275 0.002 0.000000 na
ATOM 14 C6 MOL 1 2.878 -1.373 0.005 0.000000 c3
ATOM 15 H4 MOL 1 3.505 -1.324 0.893 0.000000 h1
ATOM 16 H5 MOL 1 2.301 -2.294 0.006 0.000000 h1
ATOM 17 H6 MOL 1 3.506 -1.327 -0.883 0.000000 h1
ATOM 18 O1 MOL 1 -2.253 -2.246 -0.001 0.000000 oh
ATOM 19 H7 MOL 1 -3.150 -1.895 -0.005 0.000000 ho
To me, this seemingly confirms my suspicion in a.) This table lists
there to be 4 ca's in the molecule and one cc. However, pursuant to the
original GAFF paper, I would say that there are 3 ca's and 2 cc's in the
molecule (by my eye). Either that or I deserve an F in organic
chemistry, and hopefully Dr. Farrell is not on this list-serv.
c.) I am wondering if the antechamber algorithm for designating atom
types is meeting a pathological case? i.e., that it is assigning
incorrect atom types? I would guess that antechamber uses a
distance-based criterion for designating atom types? I am having trouble
finding out in the manual how antechamber takes xyz coordinates read in
to decide atom types?
--
Dr. Robert Molt Jr.
r.molt.chemical.physics.gmail.com
Nigel Richards Research Group
Department of Chemistry & Chemical Biology
Indiana University-Purdue University Indianapolis
LD 326
402 N. Blackford St.
Indianapolis, IN 46202
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Apr 28 2015 - 01:30:02 PDT
