Patrick Quoika

May I assume that nobody is interested in having a look at this? Or is this actually already possible, but I only couldn't figure out how? Or is this possibly...

Actually, yes. Seems to be a duplicate. As far as I understand, "textposition" in #3004 refers to what I mean with "alignment". > @Clownshift, could it be your request duplicates...

Hi Devs, I am trying to use the upper and lower walls bias modules. I wonder if there is any news about this potential issue...?

Ah okay, thanks. Yes, I understand your point. Generally, the energy does not seem to be evaluated correctly. I will check with gromacs 2022 and come back to you.

Alright, so I tried again now with gromacs-2022.5 in combination with plumed 2.9.2. I ran the example given above, with `BIASVALUE`s equal to zero. Indeed, the evaluated energy differences are...

Even though looking at the acceptance probabilities, they are significantly affected. `grep "Repl pr" npt.log` yields ``` Repl pr .96 Repl pr 1.0 Repl pr .90 Repl pr 1.0 Repl...

`-dlb no` did not fix the issue. Yes, the forcefields are identical. I use the same `topol.top`. Now for debugging, I even use the same starting conformation. Well the `dplumed`...

I run remd with this command: ``` mpirun -np 3 gmx_mpi mdrun -deffnm npt \ -multidir r{0..2} \ -dlb no -replex 500 \ -plumed plumed.dat \ -hrex ``` In this...

In the meantime, I performed some additional debugging steps. * I switched from NPT to the NVT ensemble (before, I also tried another pressure-coupling algorithm) * I use the exact...

I now tried with gromacs-2022.3 + plumed-2.8.1 and the problem still persists. Maybe I should mention that I patched gromacs with `--runtime` (otherwise I could not manage to compile +...