Multiple tc-grps, does make any sense?

User discussions
1

GROMACS version: 2023
GROMACS modification: No

Dear gmx users,
first of all, greetings, this is my first post here. I am shying entering the fascinating world of molecular simulations…
I noticed that the nvt/npt/md.mdp files report in tc-grps that, by working on two groups will led to more accurate solutions. Now, I was wondering if working with multiple tc-groups is actually the best choice anyway, or if there are any drawbacks.
Please, consider my case: I am working on a triple complex of a 50kDa enzyme bound to two different effectors in two different binding sites. I am following the usual way of action, by imposing two groups in tc-grps (Protein_Ligand1_Ligand2 Water_and_ions), but now I was wondering if here could be a better choice not to group the ligands with the protein, in this case putting “Protein Other Water_and_ions” thus considering three tc-grps. Of cource I generate the index file accordingly.
Thank you in advance
Fulvio

2

Multiple temperature coupling groups is largely a relic from ancient MD times where algorithms were inaccurate and differential heating for different groups needed to be taken care of. Nowadays I see no good reason to use multiple temperature coupling groups, except for exotic cases such as non-equilibrium MD.

3

I am wondering if we use two tc-grps for solvent and solute in a binary system with same temperature and tau_t and in another case if we use only one tc_grps for the whole system then whether there would be any problem in the first case or both will be same?

4

In practice there will not be any noticeable differences. But why use two groups when you can use one?