Ivy Zhang

@jchodera : Not sure. We may want to keep parts of this PR, as it allows for protein mutation atom mappings to be generated with the `AtomMapper` class, but we...

>I thought we wanted to keep intra-REST regions scaled by beta/beta0, but REST-environment sqrt(beta/beta0) This is what the original REST2 paper describes, so there wouldn't be any need for this...

I've profiled this to some extent with barnase:barstar and I've found that the slowest part is running `solve_mbar_for_all_states()` on large energy matrices (which makes sense, since running for more iterations...

There is already yaml support if you want to generate protein mutation input files for fah (using `perses-fah`). Here is an example: https://github.com/choderalab/perses/blob/main/perses/data/barstar-mutation/mutant.yaml This example actually hasn't been completely updated...

@jchodera @ijpulidos : Here's the log from the last 2 iterations of running ALA->ARG (charge changing): ``` ******************************* DEBUG:openmmtools.multistate.multistatesampler:Iteration 4999/5000 DEBUG:openmmtools.multistate.multistatesampler:******************************************************************************** DEBUG:mpiplus.mpiplus:Single node: executing DEBUG:openmmtools.multistate.replicaexchange:Mixing replicas... DEBUG:openmmtools.utils:Mixing of replicas took...

I just analyzed the dipeptide charge changing test data: https://github.com/choderalab/perses/issues/1044#issuecomment-1167913833

@ijpulidos : This should speed the test up a lot: ```python move = mcmc.LangevinDynamicsMove(timestep=4.0 * unit.femtoseconds, collision_rate=1.0 / unit.picosecond, n_steps=50, reassign_velocities=False, n_restart_attempts=20) ``` Note that we want to wait until...

Great! Also, I just checked and the [current tests](https://github.com/choderalab/perses/blob/main/perses/tests/test_repex.py) use 3000 iterations -- Why are you using 1000 iterations? Is the test still passing (i.e. `assert DDG < 6 *...

Ah that comment was referring to the charge changing mutations. For the neutral mutations, you'll have to see my comment [here](https://github.com/choderalab/perses/issues/1044#issuecomment-1165575916)

But both of those comments are for experiments using 50 steps per iteration, instead of 250 steps (which is what we normally use in our production runs). Are you using...