Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
| @@ -0,0 +1,1153 @@ | |||
| { | |||
There was a problem hiding this comment.
Line #25. settings=RelativeHybridTopologyProtocol._adaptive_settings(
Do we not need to add the lipid force field to the settings or is adaptive doing this? Also this is another place we are exposing _adaptive to the world do we need to add a public entry point?
Reply via ReviewNB
There was a problem hiding this comment.
i currently added the lipid forcefields in the _adaptive_settings as well, but we can think about changing that later.
| @@ -0,0 +1,1153 @@ | |||
| { | |||
There was a problem hiding this comment.
Line #43. network = openfe.AlchemicalNetwork(transformations)
Woow so many warnings we need to remove a lot of these!
Reply via ReviewNB
There was a problem hiding this comment.
Already done - Hannah's branch isn't up to date with main.
| @@ -0,0 +1,1153 @@ | |||
| { | |||
There was a problem hiding this comment.
Two out of three of these imports warning seems to be some kind of problem with our typing vendor in gufe. I have opened an issue.
Reply via ReviewNB
| @@ -0,0 +1,1153 @@ | |||
| { | |||
There was a problem hiding this comment.
For this example, rather than separating this in a separate step, I think it will be more useful to the readers if you just do these calls in-place, i.e. modeller = openmm.app.Modeller(protein.to_openmm_topology(), protein.to_openmm_positions().
With a little bit of formatting, that will look a lot less overwhelming.
Reply via ReviewNB
| @@ -0,0 +1,1153 @@ | |||
| { | |||
There was a problem hiding this comment.
This is too unweidly for users in my opinion, I think OpenFreeEnergy/gufe#568 will need to be completed.
Reply via ReviewNB
There was a problem hiding this comment.
That would be great!
| @@ -0,0 +1,1153 @@ | |||
| { | |||
There was a problem hiding this comment.
Generally would not recommend overwriting an existing file in your example, that's going to lead users to do bad things on their end - I would instead suggest you set a new file name for it.
Reply via ReviewNB
There was a problem hiding this comment.
db_complex = 'a2a/a2a_complex.pdb' is supposed to be a new file! I will remove it from the PR.
| @@ -0,0 +1,1153 @@ | |||
| { | |||
There was a problem hiding this comment.
Could you somewhere make a statement about which of these would be applicable to the "merged PDB" example shown above? (give that most Maestro users would be going down that path).
numpy array using OpenFF units (see below)
This is the wrong level of details at this point, here I would just say "box vectors in OpenMM format". Then in the section below, I would precise that it has to be a united array.
Reply via ReviewNB
There was a problem hiding this comment.
Added a maestro note and changed the box vector desciption.
| @@ -0,0 +1,1153 @@ | |||
| { | |||
There was a problem hiding this comment.
Add some details on how this is done? Any caveats that one might need to consider (i.e. making sure that the system is wrapped / not broken across periodic images)?
Reply via ReviewNB
There was a problem hiding this comment.
I added a comment on how it calculates the box and also a caveat.
| @@ -0,0 +1,1153 @@ | |||
| { | |||
There was a problem hiding this comment.
Add some details on the types of lipid forcefields that are currently supported (i.e. I know there are some that won't work, which currently do work and which should one avoid?)
Reply via ReviewNB
There was a problem hiding this comment.
Added a comment here. I haven't tried out ff other than lipid17, also commented on that.
| @@ -0,0 +1,1152 @@ | |||
| { | |||
There was a problem hiding this comment.
Do we need to say anything about how we expect the membrane to be aligned with the coordinate system so that the barostat applies the pressure correctly?
Reply via ReviewNB
There was a problem hiding this comment.
Yes, added something!
There was a problem hiding this comment.
ENH: it would be better if we move the "needs" to warning boxes and the likes. That way folks can easily see "o ok I have to be aware of this"
| @@ -0,0 +1,1152 @@ | |||
| { | |||
There was a problem hiding this comment.
Following up from Monday's meeting - please remove the CHARMM36 bit here - we don't support non-AMBER force fields.
Reply via ReviewNB
There was a problem hiding this comment.
Removed this!
| @@ -0,0 +1,1141 @@ | |||
| { | |||
There was a problem hiding this comment.
There was a problem hiding this comment.
This was just left from some trouble shooting, removed this.
| @@ -0,0 +1,1141 @@ | |||
| { | |||
There was a problem hiding this comment.
There was a problem hiding this comment.
Ah yeah, this was left over from some testing, removed this.
| @@ -0,0 +1,1141 @@ | |||
| { | |||
There was a problem hiding this comment.
This could be the entry point to OpenFE for some users so lets help them and guide them to the CLI docs for more info maybe to here https://docs.openfree.energy/en/latest/tutorials/rbfe_cli_tutorial.html#run-the-simulations ?
Reply via ReviewNB
jthorton
left a comment
There was a problem hiding this comment.
Just the two cells to remove and a link to the docs then please merge this looks great!
IAlibay
left a comment
There was a problem hiding this comment.
Can't merge as-is because the environment file is pointing to unreleased versions of openfe and gufe.
| @@ -0,0 +1,1065 @@ | |||
| { | |||
There was a problem hiding this comment.
"It is essential that the lipid molecules include bond information in the CONECT records of the PDB file to ensure proper topology recognition by OpenMM."
This sentence seems to be a bit out of the blue and/or unexplained.
I.e.
Point 1: maestro does bad things, so you might need to save the different parts separately. -- makes sense!
Point 2: It's essential that you have all the CONECT records -- is this separate from the previous point? how do I achieve this as a Maestro user?
Point 3: you can merge the components this way (follows from point 1)
I would say either remove point 2, or make it clear how it's related or move it up and add a section higher up that goes "Reading in a PDB with a membrane" and then list the caveats there?
Reply via ReviewNB
| @@ -0,0 +1,1065 @@ | |||
| { | |||
There was a problem hiding this comment.
[nit] the sentence technically reads as "OpenMM requires vectors numpy arrays with OpenFF units in their reduced form" (which they don't, just the latter).
I propose: "Box vectors must be provided as a NumPy array with OpenFF units. The vectors must be specified in their reduced form, as required by OpenMM."
Reply via ReviewNB
| @@ -0,0 +1,1065 @@ | |||
| { | |||
There was a problem hiding this comment.
The header level for validation seems to be too low. I think it should be the same size as the "loading" header font?
Reply via ReviewNB
|
The "notes" sections could be a little bit more visually striking if we used error and warning boxes. I'll open a PR targetting this one with some of these changes. |
There was a problem hiding this comment.
I just realised that this is the only tutorial under openmm_rbfe, which is a bit confusing because it's using the hybrid top protocol but RBFE could really mean septop (I know that's now how things are laid out inside of openfe, but that's also an issue in itself 😅 ).
How about moving it to just membranes for now?
No description provided.