Kyle Beauchamp

So I already have a short benchmark script (https://github.com/choderalab/pymbar/blob/master/scripts/benchmark_mbar.py). IMHO, we don't want to invest more time in automated performance testing given the huge list of other tasks. Small changes...

The `nk` branch uses subsampling (along the `N` axis) based guessing, which sidesteps this issue.

This was implemented in several of my performance based pull requests, but I will defintely need to make another pass of benchmarking and cleanup. I'll re-open this issue as a...

Per discussion, we have decided: 1. For 2.0, we make minimal changes necessary to have useful docs (probably readthedocs.org) 2. For 3.0, we make more full-featured docs in the style...

No, we still need to figure out how we want to divide things between: 1. readthedocs 2. README.md 3. Example python scripts

I had a similar idea: for any cell that's denoted as "code" I'd like the default to be output only. Basically, I want any option that by default hides all...

So it looks like ProPKA won't actually do the protonation--it just spits out the pKa which you can use in PDB2PQR to adjust the residue. However, it might be possible...

So after more digging, I've uncovered two major flaws with the idea of using PROPKA for this purpose. First, as mentioned above, ProPKA doesn't actually spit out a PDB structure....

H++ is probably difficult to get running locally--they don't have installation files available except via personal communication. MCCE might be a good idea.

PS: we should get MMTools on GitHub at some point.