John Chodera

I'm afraid the examples have drifted out of date as the YAML syntax was revised. We're shortstaffed due to COVID, but hope to have more folks coming aboard to remediate...

The [yank recipe](https://github.com/omnia-md/conda-recipes/blob/master/yank/meta.yaml#L29) requires [`mpiplus`](http://github.com/choderalab/mpiplus). We migrated the recipe to conda-forge, but it appears we [neglected to include that dependency](https://github.com/conda-forge/mpiplus-feedstock/blob/master/recipe/meta.yaml). Could you refile the issue that mpiplus should require mpi4py...

@andrrizzi : Should we pin YANK's recipe to `pymbar==3.0.3` until @mrshirts figures out who broke pymbar and why?

If we wanted to stick with a Python writer we could still use [MDTraj](http://mdtraj.org) to write an XTC file and binary concatenate them.

@andrrizzi : Were you able to address this in an `openmmtools` update?

It _should_ be relatively straightforward to add a simple additional step to the pipeline to re-assign parameters to the whole system using the [`openmmforcefields` `SystemGenerator`](https://github.com/openmm/openmmforcefields/#automating-force-field-management-with-systemgenerator). I'll make a plan to...

@daveminh : Can you post a tarball of some input files you would like to use through the normal YANK pipeline? If you configure them with GAFF now, I can...

Hm. Can you give some examples that fail? Maybe we can come up with a pipeline that works around this.

Here's [Figure 3 from the Jin et al. paper](https://www.biorxiv.org/content/10.1101/2020.02.26.964882v2):  I don't think openff can cover Ebselen (which has a selenium), but I think we can cover the rest. I'll...

The [log link](https://github.com/daveminh/Chem456/tree/master/static_files/tutorials/3cl-pro/YANK/output/ZINC000001714738/experiments/nan-error-logs) you posted doesn't seem to resolve.