Hi, Chris,
AM1-BCC method does corrections (based on both atom types and bond
types) on AM1 Mulliken charges. The difference between the two sets of
charges was caused by the difference of the AM1 Mulliken charges of the
two conformations. Please refer to the attached output of AM1
optimization by mopac. It is said that AM1 Mulliken charges are not very
sensitive to conformations (at least compared to RESP charges), and this
feature is considered as one of the advantages of AM1-BCC method.
However, for this particular case, the differences is very large. I am
not clear what is the reason.
All the best
Junmei
Dear Amber(l)ist ,
I found that charges calculated with Antechamber/mopac (v1.25) on two
very close ligand conformation (RMSD = 0.45A) could lead to some
surprising different charges on Atoms.
antechamber -nc 5 -rn lig -i liga_regul.pdb -fi pdb -o liga_regul.mol2
-fo mol2 -c bcc -j 4 -at gaff -pf y
antechamber -nc 5 -rn lig -i lig_mini.pdb -fi pdb -o lig_mini.mol2 -fo
mol2 -c bcc -j 4 -at gaff -pf y
Please find in the attachment the mol2 files "liga_regul.mol2" and
"lig_mini.mol2" , H7 (from NH3+) has a charge of +0.05 in
"liga_regul.mol2" where is it +0.48 in "lig_mini.mol2".
Graphicaly looking to these conformations, I can't understand why H7
charge is so different , and moreover its charge is different than the
two other H in -NH3+ (HN51, HN52)
It looks like it is structure dependent, if you start again from the
previous mol2 files , you have the same result and charge different for
H7. antechamber -nc 5 -rn lig -i liga_regul.mol2 -fi mol2 -o
liga_regul_NEW.mol2 -fo mol2 -c bcc -j 4 -at gaff -pf y antechamber -nc
5 -rn lig -i lig_mini.mol2 -fi mol2 -o
lig_mini_NEW.mol2 -fo mol2 -c bcc -j 4 -at gaff -pf y
My question is the following:
1) It is normal to find such charges difference for a slightly
ddifferent conformation ?? If so , I guess, it would be a good idea to
first optimize (I thought MOPAC would do it) the conformation before
calculating the charges with Antechamber.
Thanks a lot for any thought.
Chris
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Aug 27 2006 - 06:07:20 PDT
