Unrealistic qeq and qtpie charges

Previous Topic Next Topic
classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

Unrealistic qeq and qtpie charges

I am trying to use openbabel (version 2.4.1) to calculate approximate
charges using qtpie and qeq, and the results that I obtain appear
completely unreasonable. For example, using the following .xyz file for
water (O-H = 0.947 A, HOH = 105.5):

O      0.0000    0.0000    0.0000
H      0.7538    0.5732    0.0000
H     -0.7538    0.5732    0.0000

I obtain a qtpie charge of -1.5083 and a qeq charge of +1.8296 on the
oxygen atom. Other charge models give values much more in line with
expectations (eem(qO) = -0.6501, gasteiger(qO) = -0.4105, and mmff94(qO)
= -0.8600).

The following command was used to determine charges:
      obabel water.xyz -omol2 --partialcharge qtpie

Thinking that this might simply due to the fact that the water molecule
is not modeled well by these approximate methods, I looked at phenol
(this was molecule was included in the original qtpie paper (Chen,
Martinez, J. Chem. Phys. Lett., 2007, 315)). Again, unreasonable charges
were obtained the qeq and qtpie charges were obtained. For example,
using qtpie the calculated charge on the O was -1.4532 and the charge on
the aromatic carbon bound to the OH group was +2.0941.

I obtained the same results using the precompiled version 2.3.2 of
openbabel included with Linux Mint on a different computer. I did note
that the source code for the qtpie routine indicates that some changes
were made from the published procedure, so the values obtaine would not
be identical. However, this note also indicated that very similar values
should be obtained.

Am I the only person having this problem, or is there a error in the
implementation of the qeq and qtpie charge models?

Thank you.

Developer Access Program for Intel Xeon Phi Processors
Access to Intel Xeon Phi processor-based developer platforms.
With one year of Intel Parallel Studio XE.
Training and support from Colfax.
Order your platform today.http://sdm.link/xeonphi
OpenBabel-discuss mailing list
[hidden email]