REST2 simulations not exchanging

[Theodore Fobe]

Hello Plumed community!

These forum posts have been extremely helpful with getting Plumed working in the past, so thank you to everyone for contributing to this thriving community of computational scientists!

I'm running into an issue getting HREX working for a protein simulation. I'm working with Gromacs 2024.2 patched with Plumed  2.9.3. I was able to generate scaled topologies for my system using the Hamiltonian Replica Exchange tutorial in the documentations pages, however when I try to run simulations with these topologies, the simulations do not exchange at all. Here is an example exchange output for an 8 replica system that was started with this call to gromacs:

mpirun -np 8 gmx_mpi mdrun -v -multidir replica_{0..7} -deffnm nvt_short -replex 100 -ntomp 1 -plumed plumed.dat -hrex 

Replica exchange at step 322900 time 1291.59998

Repl 0 <-> 1  dE_term = -0.000e+00 (kT)

dplumed =  5.502e+05  dE_Term =  5.502e+05 (kT)

Repl ex  0    1    2    3    4    5    6    7

Repl pr   .00       .00       .00       .00

I did go through the checks mentioned in the tutorial about running two simulations with the same topologies and I get the same result, where there is no exchange between two simulations with the same topology. For each of these swaps the `dplumed` energy difference is always has a magnitude of 1e5, even for two simulations run with the same topology. I don't know why the `dplumed` energy is so large? Is there a way to print the contributions to this term?

Some other details about my system, I get the same result using both ff19sb and OPC  water and ff14sb  and TIP3P water. For the ff19SB system I manually scaled the CMAP terms following recommendations from other posts on this forum. However, since both of these force fields give the same issue of a very large `dplumed` value, I don't think it's a force field problem.

I've attached my original topology and a scaled topology, an equilibrated starting structure and my MDP file used to generate these results.

Any guidance on this would be very appreciated! Thanks

Lenny

[Giovanni Bussi]

Hi, hrex is not yet supported in the 2024 patch

Sent from Gmail mobile

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/plumed-users/30cd3ee7-e85d-4f5c-861a-9b8a1727e120n%40googlegroups.com.

[Theodore Fobe]

Hi Giovanni,

I retried these simulations using Gromacs 2023.5 and was able to get a much lower `dplumed` value. I'm still adjusting parameters to get good exchange rates, but the sanity check of running HREX with two of the same topologies gave an exchange probability of 1 and `dplumed` values close to zero, so now it may just be a matter of getting good energetic overlap between replicas. To that end, I've been looking at EDR file energy outputs. For my current setup when running 16 simulations over the temperature range of 300 K to 500 K, I see system potential energy overlap, however I'm still getting poor exchanges (~4-6% exchange probability).

Here is a sample RE output from the simulations that produced the potential energy plot above:

Replica exchange at step 1935100 time 7740.39990

Repl 0 <-> 1  dE_term = -0.000e+00 (kT)

dplumed =  1.754e+00  dE_Term =  1.754e+00 (kT)

Repl ex  0    1    2    3    4    5    6    7    8    9   10   11   12   13   14   15

Repl pr   .17       .00       .00       .01       .00       .00       .00       .00

 Now `dplumed` values are more reasonable on the order of magnitude of 1e-1 to 1e+1. What I find odd here is that despite the good overlap in potential energy I still get fairly low exchange probabilities. Would you have any recommendations on getting better exchange probabilities? The enzyme I'm using partial tempering on has 5767 atoms, but I don't have a heuristic for what the `dplumed` values should be for a system with that many atoms.

Thanks again for your help with this!

Lenny Fobe

You received this message because you are subscribed to a topic in the Google Groups "PLUMED users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/plumed-users/fEJ8YuVwCP4/unsubscribe.
To unsubscribe from this group and all its topics, send an email to plumed-users...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/plumed-users/CAPLm8ZK%3DinGOs2oOv9Yx-F%2BHZoAyq%2B-td6mg_g3BWrXmMmkA%3Dw%40mail.gmail.com.