C-rescale vs Parrinello–Rahman in final NPT equilibration before protein–protein SMD (TCR–pMHC, ~1M

User discussions
1

GROMACS version:
GROMACS modification: Yes/No
Here post your question

Hi everyone,

I’m running GROMACS simulations for a protein–protein system (TCR–pMHC complex) and planning to perform steered MD (SMD) after equilibration. I have a large system with approximately 1 million atoms, and I am trying to gradually release the position restraint over multiple stages in NPT equilibration.

My equilibration protocol is:

image
image791×429 10.5 KB

Is it good practice to switch from C-rescale (first-order) to Parrinello–Rahman (second-order) in the final NPT stage before production?

Would it be better to keep C-rescale throughout, including the final unrestrained stage?

For a large protein–protein system like this, are tau_p = 3.0 for restrained NPT and tau_p = 2.0 for the final unrestrained NPT reasonable choices?

I have read somewhere about switching to first order barsotat as those correctly reproduce the correct statistical ensemble.

Would appreciate any advice or thoughts on this.

Thank you !

2

C-rescale is the barostat of choice. This produces the correct ensemble and does not have the issues with oscillations and instabilities of PR. I would suggest=5 ps. tau_p will also produce correct results, but the performance might be affected slightly.