VMD-L Mailing List

From: Mayne, Christopher G (cmayne2_at_illinois.edu)
Date: Sat Jan 20 2018 - 20:05:43 CST

Ernest,

The purpose of these calculations is to find the distance and angle of water interaction at the optimized against energy.

For the cases in which the water molecule moves far from the interaction site, there is no local minimum near typical interaction distances; therefore, this data point is of no value to the charge optimization — leave it out.

For the cases where the optimization does not converge, try adjusting the optimization convergence criteria by adding Opt=Loose to the Route. See Gaussian’s documentation for further information.

For the cases where the interatomic distances are causing errors, check that the input structures generated by ffTK in the GAU files are reasonable. This error suggests that the molecule has 3D architecture and that the placed water clashes with another portion of the molecule. When this occurs for interaction sites on planar substructures, e.g., aromatic rings, ffTK has arbitrarily selected one side of the ring vs the other. The quick fix is to manually edit the Z-matrix of the GAU input file to flip the water molecule to the other side of the ring. It is also worth noting, that if the clash occurs with a portion of the molecule that is far away (by connectivity) from the water interaction site, you should consider designing smaller model fragments to parametrize — this will save you huge amounts of time when computing more rigorous target data such as the hessian for bond/angle optimization and torsion scans for dihedral optimization.

Regards,
Christopher Mayne

> On Jan 20, 2018, at 5:09 AM, Ern Ong <Ern.Ong_at_student.adfa.edu.au> wrote:
>
> Hi VMD users,
>
> I encountered error terminations at Gaussian09 for 2 out of 12 HB acceptors and 3 out of 13 HB donors. The Gaussian input files for water interaction were generated using the default settings: # HF/6-31G* Opt=(Z-matrix,MaxCycles=100) Geom=PrintInputOrient (VMD1.9.3 2016-11-30 Windows platform). My molecule of interest is a single negatively charged mannose with a pyruvic acid group. The following is the summary of error messages:
>
> #1 HB acceptor:
> Item Value Threshold Converged?
> Maximum Force 0.000698 0.000450 NO
> RMS Force 0.000493 0.000300 NO
> Maximum Displacement 0.000349 0.001800 YES
> RMS Displacement 0.000247 0.001200 YES
> Predicted change in Energy=-2.434687D-07
> Optimization stopped.
> -- Number of steps exceeded, NStep= 100
> -- Flag reset to prevent archiving.
> Error termination request processed by link 9999.
>
> #2 HB acceptor:
> Item Value Threshold Converged?
> Maximum Force 0.000650 0.000450 NO
> RMS Force 0.000461 0.000300 NO
> Maximum Displacement 0.004805 0.001800 NO
> RMS Displacement 0.003399 0.001200 NO
> Predicted change in Energy=-1.559172D-06
> Optimization stopped.
> -- Number of steps exceeded, NStep= 100
> -- Flag reset to prevent archiving.
> Error termination request processed by link 9999.
>
> #1 HB donor: (This water molecule is “spinning and flying” away from the very beginning)
> Item Value Threshold Converged?
> Maximum Force 0.000058 0.000450 YES
> RMS Force 0.000041 0.000300 YES
> Maximum Displacement 1.403879 0.001800 NO
> RMS Displacement 1.011110 0.001200 NO
> Predicted change in Energy=-2.495560D-05
> Optimization stopped.
> -- Number of steps exceeded, NStep= 100
> -- Flag reset to prevent archiving.
> Error termination request processed by link 9999.
>
> #2 HB donor: (This water molecule is also “spinning and flying” away from the very beginning)
> Item Value Threshold Converged?
> Maximum Force 0.000013 0.000450 YES
> RMS Force 0.000009 0.000300 YES
> Maximum Displacement 0.489277 0.001800 NO
> RMS Displacement 0.424264 0.001200 NO
> Predicted change in Energy=-2.355976D-05
> Optimization stopped.
> -- Number of steps exceeded, NStep= 100
> -- Flag reset to prevent archiving.
> Error termination request processed by link 9999.
>
> #3 HB donor: (Since the water molecule is automatically assigned linearly at a 2 angstrom distance, the water molecule is found to be overly close to other atoms)
> Small interatomic distances encountered: 33 29
> Problem with the distance matrix.
> Error termination via Lnk1e in
>
> Hope that someone can help me solving the problems. I am not sure with the accuracy of charge optimization if I omit these water molecules, since I do not wish to further optimize the charges of atoms that these water molecules interact with.
>
> Thank you.
>
> Best Regards,
>
> Ernest Ong Ern Seang
>
> PhD student
> School of Engineering and Information Technology (SEIT)
> University of New South Wales@ ADFA
> NorthCott Drive, Campbell
> ACT 2600, Australia.


  • application/pkcs7-signature attachment: smime.p7s