Dear Dr Case, dear Crystal.CHIA-YU Ku,
I looked at this thiamine mono-phosphate structure:
http://fr.wikipedia.org/wiki/Thiamine_pyrophosphate
- first, I am surprised by chemical equivalencing generated by
Antechamber in this "tps_am1bcc.mol2" molecule.
I think the charges of the three oxygens should be equivalenced:
20 O1 30.3800 41.0610 66.6910 o 97 TPS -0.8618
21 O2 31.2660 38.8010 66.1190 o 97 TPS -0.8045
22 O3 30.3310 39.2850 68.3790 o 97 TPS -0.4925
same remarks for the methyl hydrogens below
23 HM21 16.7330 40.3920 63.8210 hc 97 TPS 0.0572
24 HM22 16.9320 39.0060 64.9200 hc 97 TPS 0.0613
25 HM23 17.6920 38.9820 63.3110 hc 97 TPS 0.0628
27 HM41 25.5760 40.6420 64.4370 hc 97 TPS 0.0754
28 HM42 24.7600 42.1100 65.0270 hc 97 TPS 0.0203
29 HM43 23.8280 40.8530 64.1770 hc 97 TPS 0.0198
- second, optimizing in gas phase a molecule that contains a +1 formal
charge and a -2 formal charge in two different parts should lead to a
non-representative conformation.
See
http://q4md-forcefieldtools.org/REDS/faq.php#21 to share data
between users
-> These two points are clearly exemplified by a R.E.D. Server job available .
http://cluster.q4md-forcefieldtools.org/~ucpublic1/ADF1ADF-Thiamine-phosphate/P4692.html
- the P2N input file reflecting correct chemical equivalencing is at:
http://cluster.q4md-forcefieldtools.org/~ucpublic1/ADF1ADF-Thiamine-phosphate/P4692/Mol_red1.p2n
- the bad geometry obtained after geometry optimization is demonstrated .
http://cluster.q4md-forcefieldtools.org/~ucpublic1/ADF1ADF-Thiamine-phosphate/P4692/javaappletpdb-1.html
- the correct chemical equivalencing for this bad geometry for the
methyl groups & 3 oxygen atoms of OPO3 is at
http://cluster.q4md-forcefieldtools.org/~fydred/Project/P4692/Data-R.E.D.Server/Mol_m1-o1.mol2
1 CM2 6.082805 0.072160 0.815600 C 1 TPS -0.2715
2 HM21 6.627597 0.818481 0.247071 H 1 TPS 0.0608
3 HM22 6.595395 -0.877743 0.751613 H 1 TPS 0.0608
4 HM23 6.062396 0.404810 1.848071 H 1 TPS 0.0608
24 CM4 0.047228 -0.189347 1.714181 C 1 TPS -0.1828
25 HM41 -0.474275 -0.242833 2.657478 H 1 TPS 0.1004
26 HM42 0.052174 0.841373 1.392977 H 1 TPS 0.1004
27 HM43 1.055407 -0.559694 1.859550 H 1 TPS 0.1004
35 P1 -2.184486 1.683740 -0.497671 P 1 TPS 0.9763
36 O1 -0.738195 2.109109 -0.265858 O 1 TPS -0.7681
37 O2 -3.165967 2.757597 -0.774316 O 1 TPS -0.7681
38 O3 -2.267562 0.434272 -1.347688 O 1 TPS -0.7681
My feeling is that using a building block approach based on the
thiamine and methylphosphate [Me-OPO3(2-) and Me-OPO3H(-)] molecules
through R.E.D. Server will lead to (i) correct chemical equivalencing
(ii) representative conformations (iii) charge values for thiamine,
thiamine monophosphate_s and thiamine pyrophosphate_s and even
more..., and to (iv) obviously reproducible charge values...
regards, Francois
Quoting David A Case <case.biomaps.rutgers.edu>:
> Message from Junmei Wang below clears up two problems in this thread:
>
> 1. The reason acdoctor failed was because the input pdb was not formatted
> correctly. (Perhaps acdoctor should have diagnosed this problem as well,
> but once things get in the wrong columns in PDB files, it is very difficult
> to un-do the damage.)
>
> 2. Here is another case where mopac converges, whereas sqm and Gaussian 09
> did not. Afficianadoes of sqm convergence problems may want to see if they
> can figure out why.
>
> Thanks to Junmei for the help.....dac
>
> ----- Forwarded message from Junmei Wang <junmwang.yahoo.com> -----
>
> Date: Thu, 17 Nov 2011 07:46:22 -0800 (PST)
> From: Junmei Wang <junmwang.yahoo.com>
>
> I found the lines from "ATOMÂ Â Â 10" to "ATOMÂ Â 38" shift one
> column. This is probably the reason leads to acdoctor fail. pdb format is
> very tricky and the fields must be reside in certain columns.
>
>> From the graphics, I can see the net charge of this molecule is -1.
>
> I ran the following command to assign am1bcc charges:
>
> antechamber -fi pdb -fo mol2 -i tps.pdb -o tps_am1bcc.mol2 -nc -1 -c bcc
>
> This command failed since sqm cannot converge.
>
>
> Then I replace sqm with mopac and the command can be run smoothly:
>
> antechamber -fi pdb -fo mol2 -i
> tps.pdb -o tps_am1bcc.mol2 -nc -1 -c bcc -df 0
>
> Please find the attached tps_am1bcc.mol2 file
>
> Junmei
> ----- End forwarded message -----
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Nov 18 2011 - 09:00:03 PST
