[gmx-users] g_dipoles + tpbconv

[Justin A. Lemkul]

Nilesh Dhumal wrote:
> Hello,
> 
> I have a system with a glucose molecule.
> 
> I want to calculate the dipole moment of a particular OH in glucose
> molecule.I made an index file which have a group containing atoms.
> 
> [ O8 ]
>   8   18
> 
> 10 is oxygen no. and 20 is hydroen no.
> 
> 

I'm assuming you mean 8 is O and 18 is H?

> 
> But, if I try to use g_dipoles to get the dipole moment contribution from 
> using following command:
> 
> g_dipoles -f 6.trr -s 6.tpr  -n index.ndx -corr mol -nonormalize  -c
> 
> I get the following error.
> 
> Fatal error:
> index[1]=8 does not correspond to the first atom of a molecule
> 
> I read on gmx-users mailing list.
> 
> http://www.mail-archive.com/gmx-users@gromacs.org/msg40091.html
> 
> Calculating a dipole moment of a possibly-charged species requires that
> there be a reference point, which is conventionally the center of mass of
> the molecule. This means all the atoms of the molecule have to be known,
> and g_dipoles assumes the index group consists of whole molecules.
> 
> So to do the partition you want, you will need to provide the same
> molecules, but with zero charges. That will mean making a copy of your
> .top and hacking those charges to zero to generate a new such .tpr.
> 
> Actually, tpbconv has the ability to set the charges of a group in a .tpr
> to zero.
> 
> So I tried
> 
> tpbconv -f 6.trr -s 6.tpr -n glu-emi-cl-128-no.ndx -e 6.edr  -o zero.tpr
> 

This command doesn't accomplish anything.  If you're trying to zero out the 
charges, you need to use the -zeroq option.  But then, if you have two atoms 
with zero charge, then the dipole will of course be zero!

> I got the error
> 
> segmentataion fault
> 
> How can I solve this problem to calcualte the dipole moment of a bond?
> 

I doubt you can do it with g_dipoles.  You've been told already that the group 
analyzed has to have a net charge of zero.  An OH group will not satisfy that 
need, nor does assigning each of the atoms zero charge.  You're better off using 
g_dist (or maybe g_bond) to find the distance between the O and H atoms over 
time and plug that into an equation using the partial charges on the atoms to 
calculate the dipole yourself.  The charges are fixed, so the only thing that 
changes (in the absence of constraints) is the distance between them.

-Justin

> I am using gromacs version 4.0.7.
> 
> Thanks
> 
> Nilesh
> 
> 
> 
> 
> 
> 

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================