[gmx-users] Gromacs 5.1 insert-molecules - Generating bad contacts?

[Dan Gil]

Hi,

I am upgrading from gromacs 4.6 to gromacs 5.1. Gromacs 4.6 had "genbox,"
which I was able to use without a problem.

With gromacs 5.1, I am using insert-molecules to generate my initial
configuration containing water, ethanol, and a lennard-jones particle which
I defined. When I attempt to minimize the energy using "steep", the program
crashes, apparently due to clashes.

Here is an excerpt of the error right before the program crashes:

Step 19, time 0.038 (ps)  LINCS WARNING
relative constraint deviation after LINCS:
rms nan, max 0.005299 (between atoms 9321 and 9317)

-----

Also another side note, I was able to run the program on a personal
computer using single-precision. When I run it with double-precision, it
crashes.

Best Regards,

Dan