The number of decimal places of atom partial charges is fixed in the GMX?

User discussions
1

GROMACS version:2020.2
GROMACS modification:No
Dear all gmx users,
I want to conduct a simulation on a polymer, which is made up of numerous residues. I wrote the atom charges in the .rtp file, but most of these value have more than 6 decimal places (such as 0.123456789 ,etc.). It will give an error in the preparation of the .tpr file in the energy minimization which said the net charge is not an integer. I found that this is because the atom charges in the .itp file (produced by pdb2gmx) are rounded due to the number of decimal places is fixed at 6. Is there any solution to solve this? I want to expand the scope of number of decimal places, which code file should be modified? Maybe this is a little question, but great thanks!

Best Regards
Little Ma

2

Are you intending to run the simulations in double precision? If not, you’re never going to benefit from such extreme precision. The limiting factor will be the precision of the coordinates. Most force fields don’t go beyond 5 decimal places (AMBER), while others rely on far fewer. I would be very skeptical of how useful all those extra digits really are.

3

Thank you very much for the kind reminder, which help me a lot.
I will round the partial charges to a suitable values. Thank you again! (^_^)
@jalemkul