Add tutorial for running REST #17
Conversation
There was a problem hiding this comment.
This looks really nice!
It's quite a complicated process, involving a lot of code. How about adding a few paragraphs at the beginning to give an overview of the process? Explain what steps you're going to perform and why. That will provide some context to help them understand each step as they come to it.
I found the get_rest_identifier() function confusing, because when you come to it you don't yet understand what it's for. Why are you creating lists of 1s and 0s to represent the different types? The answer is that you're setting values of parameters for custom forces, but the reader doesn't know that yet. How about moving this function down to after you create the CustomBondForce? At that point it will be more obvious why it's written the way it is.
| } | ||
| ], | ||
| "source": [ | ||
| "pdb = app.PDBFile('/home/zhangi/choderalab/openmm-cookbook/notebooks/cookbook/villin.pdb')\n", |
There was a problem hiding this comment.
Absolute paths won't work on users' computers. Can you use a relative path instead, and provide a link for the user to download the file (https://raw.githubusercontent.com/openmm/openmm-cookbook/main/notebooks/cookbook/villin.pdb)?
| "source": [ | ||
| "Replica exchange solute tempering (REST) is an enhanced sampling approach that aims to improve sampling for a user-defined subset of atoms (i.e., REST atoms) by increasing its effective temperature.\n", | ||
| "\n", | ||
| "REST involves running replica exchange to sample a set of thermodynamic states, where one state contains the unmodified Hamiltonian and all other states involve modified Hamiltonians where REST atom interactions are scaled to increased temperatures (Wang et al. 2011, DOI: 10.1021/jp204407d).\n", |
There was a problem hiding this comment.
Can you include a hyperlink to the paper?
| "\n", | ||
| "REST involves running replica exchange to sample a set of thermodynamic states, where one state contains the unmodified Hamiltonian and all other states involve modified Hamiltonians where REST atom interactions are scaled to increased temperatures (Wang et al. 2011, DOI: 10.1021/jp204407d).\n", | ||
| "\n", | ||
| "In this tutorial, we show how to build an REST-capable OpenMM system. At the end, we show how to use OpenMMTools (https://github.com/choderalab/openmmtools) objects to run REST." |
There was a problem hiding this comment.
It might be better to link to the documentation (https://openmmtools.readthedocs.io) rather than the github repo.
|
@peastman : Thanks so much for the feedback! I've addressed all your comments and pushed my changes. |
zhang-ivy
changed the title
|
If this is based on code in perses, would it make sense to show how to do this using perses? I guess the question is, how would you recommend that someone do this in practice? Would you recommend that they do everything from scratch as shown here? Or would you recommend that they use perses? Whichever approach is preferred, that's what the tutorial should show. #16 is merged, so we now have |
This tutorial is inspired by a factory in perses, but the perses factory creates REST-capable alchemical systems only, and doesn't allow for REST-capable non-alchemical systems. This tutorial shows how to create the latter, which is much simpler. I think ideally the code in this tutorial could be moved to openmmtools, but for the time being, I would recommend creating the system from scratch. |
|
Ok, sounds good. In that case, if you just move it to the tutorials folder I think it will be good to merge. |
|
Just moved the notebook (and made a copy of the required input PDB) in |
|
I think that means it's ready to merge. Thank you! |
Here's my latest draft of a tutorial for building a REST system (using OpenMM) and running a REST simulation (using openmmtools). This is a part of a larger effort to draft OpenMM 8 tutorials for the paper.
Please let me know if anything is unclear or there are any typos.
Also note that we will want to rename the notebook once it's finalized.