T019 - Molecular dynamics simulation #56

Mika-Le · 2020-10-26T12:51:07Z

Details

Talktorial ID: 019
Title: Molecular dynamics simulation
Original authors: Pietro Gerletti
Reviewer(s): Mareike Leja, David Schaller
Date of review: 09/2020-present

Content review

Potential labels or categories (e.g. machine learning, small molecules, online APIs): interaction, small molecule, protein
One line summary: Perform a molecular dynamics simulation of the SARS-CoV-2 main protease in complex with an inhibitor.
The table of contents reflects the talktorial story-line; order of #, ##, ### headers is correct
URLs are linked with meaningful words, instead of pasting the URL directly or linking words like here.
I have spell-checked the notebook
Images have enough resolution to be rendered with quality, without being too heavy.
All figures have a description
Markdown cell content is still in-line with code cell output (whenever results are discussed)
I have checked that cell outputs are not incredibly long (this applies also to DataFrames)
Formatting looks correctly on the Sphinx render (bold, italics, figure placing)

Code review

review-notebook-app · 2020-10-26T12:51:14Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

schallerdavid · 2020-11-26T21:29:24Z

Hey @jaimergp,

I could not render with sphinx easily because this notebook is probably missing in some sort of configuration file. How can I include it?

jaimergp · 2020-11-27T14:15:12Z

Hey @jaimergp,

I could not render with sphinx easily because this notebook is probably missing in some sort of configuration file. How can I include it?

Check docs/talktorials and create the correaponding nblink file there!

schallerdavid · 2020-12-03T14:02:12Z

@AndreaVolkamer Everything works well as notebook, colab notebook or html 🎉

Ready for review!

dominiquesydow · 2020-12-11T09:53:33Z

@Mika-Le - your talktorial's final index will be T019. Could you please at some point update this PR according to that (let me know if you need help)?

AndreaVolkamer · 2020-12-17T13:30:57Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


Add info to authors, see e.g. T012

Reply via ReviewNB

AndreaVolkamer · 2020-12-17T13:30:57Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


Prepare the protein ligand comples -> complex

Suggestion: I would move the 'Merge protein and ligand' one level lower:
Prepare the protein ligand complex
Prepare protein
Prepare ligand
Merge protein and ligand

Reply via ReviewNB

AndreaVolkamer · 2020-12-17T13:30:57Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


You so nicely explained what to expect in all references, despite
Pierre-Simon Laplace, Oeuvres Complètes de Laplace. Théorie Analytique des Probabilités (volume VII Gauthier-Villars (1820), 3rd ed)
Could you add sth here also?

Reply via ReviewNB

AndreaVolkamer · 2020-12-17T13:30:57Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


Maybe use a block quote for the cited text, see eg T012 (same as linked above)

Reply via ReviewNB

AndreaVolkamer · 2020-12-17T13:30:57Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


I saw in some other notebooks that they adapted the figure size, maybe one could apply this here too? check T016 figure 1

Reply via ReviewNB

I thought I remembered from our first Hackathon that we want to avoid HTML Syntax, which is the only way I know to scale images in notebooks. But I cannot find the guidelines anymore and might be mistaken. Maybe Jaime can tell us.

AndreaVolkamer · 2020-12-17T13:31:00Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


maybe also here one could comment a bit more on the solvent part

Reply via ReviewNB

AndreaVolkamer · 2020-12-17T13:31:00Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


This is starting to go into the details (and maybe too much for an introductory talktorial), but yould we at least link to where to find more information about what e.g. the 'LangevinIntegrator' does?

Reply via ReviewNB

Good Point! I added two sentences and links for further info for now.

AndreaVolkamer · 2020-12-17T13:31:00Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


set-up or set up, try to use it consistently throughout the notebook

Reply via ReviewNB

AndreaVolkamer · 2020-12-17T13:31:01Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


The comments here and in the next few cells could rather become markdown cells that guide through the process:

Minimize energy and output the minimized system
simulation.minimizeEnergy() with open(DATA / "topology.pdb", "w") as pdb_file: app.PDBFile.writeFile(simulation.topology, simulation.context.getState(getPositions=True, enforcePeriodicBox=True).getPositions(), file=pdb_file, keepIds=True)

Reply via ReviewNB

AndreaVolkamer · 2020-12-17T13:31:01Z

teachopencadd/talktorials/018_md_simulation/talktorial.ipynb

@@ -0,0 +1,1155 @@
+{


Is the number for the 'next' notebook already fixed? If yes, please include it here.
... refer to Txxx ... (maybe even with a link if available)

At the very and, we usually have a small quiz (three questions) can you think of any?

Reply via ReviewNB

The number is T020.
We could link to the right folder in GitHub or do we wait for the website?

AndreaVolkamer · 2020-12-17T13:37:14Z

@Mika-Le the talktorial is GREAT!!! Very easy to follow and very good explanations, well done!
I've added my very little feedback above.
MD simulations are very complex and require quite some tools/libraries, and I think the depth you are covering is great. Since - especially in the practical part some people might still want to know more about the individual steps or libraries, could you please include the links to those there again.
[Note, I did only go through it on reviewnb, without running it on colab, I'll do that as soon as time allows]

Mika-Le · 2021-01-20T15:05:56Z

Hi @AndreaVolkamer, thank you for the great feedback. I updated the notebook, it is ready to be reviewed again.

AndreaVolkamer · 2021-01-24T19:55:57Z

@Mika-Le very well done (@schallerdavid), the notebook is great! I only added tiny textual adaptions.

Unfortunately, I couldn't get it running on colab due to the mentioned GLIBCXX_3.4.26 import error. I tried some fixes from this link, but after the rdkit error was fixed, I got a No module named 'mdtraj.formats.dcd' error?
Maybe @jaimergp since you looked into colab for the other course you have a fix in hand?

As soon as the colab thing is solved, the notebook is ready to be merged!

jaimergp · 2021-01-25T10:51:55Z

As per our chat, the issue is solved but we can't make sure it works reliably right now because Anaconda.org servers are a bit irresponsive now and installation fails. We have decided to reconvene in a week, since by then openmmforcefields might be on conda forge already (and does not suffer from server issues since it runs on separate network).

jaimergp · 2021-02-02T13:25:01Z

openmmforcefields not yet on conda-forge due to blocking openmm/openmmforcefields#151. We'll wait a bit so we can merge with a conda-forge only env!

schallerdavid · 2021-02-12T12:55:18Z

@jaimergp,
everything works out with installing only via conda-forge 🎉. I would be done from my side.

jaimergp · 2021-02-12T14:52:09Z

teachopencadd/talktorials/019_md_simulation/README.md

+
+## Categories
+
+This talktorial is part of the following categories: [Collections overview](link)


I think this is outdated. Are we writing the categories in the notebook now?

jaimergp · 2021-02-12T15:12:42Z

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

@@ -0,0 +1,1250 @@
+{


Note that !command does not stop the cell from being executed. "Dependencies successfully installed" will be showed regardless the success of the conda install command. This, in addition to the cleared outputs, can be very misleading... We should check whether things were actually installed or not, maybe try importing openmm or something similar. Only if that succeeds we clear outputs and report success. Otherwise we want the logs to see what went wrong (e.g. the random network errors we saw).

Reply via ReviewNB

Is it pythonic/ok to nest try except blocks? Something along the lines of
try: import condacolab [...] try: import rdkit clear_output() print("Installation successful") except: print("Dependencies not installed") except: print("Not on Colab")

If not, what would be the better way to check for successful installation? Any If-else-checks possible for installed dependencies?

I'd use the else clause of the try block, which will only be reached if no exception was raised during try

try: import condacolab ... except: print("Not on Colab") else: try: import rdkit clear_output() print("Installation successful") except: print("Dependencies not installed!")

Thanks, that is way better

jaimergp · 2021-02-12T15:12:42Z

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

@@ -0,0 +1,1250 @@
+{


We capitalize Python and Protein Data Bank.

Reply via ReviewNB

jaimergp · 2021-02-12T15:12:42Z

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

@@ -0,0 +1,1250 @@
+{


RCSB offers a simple way to get PDBs via URLs: https://files.rcsb.org/download/3POZ.pdb. Do we need the get_pdb_file function? We could use requests instead. Up to you!

Reply via ReviewNB

jaimergp · 2021-02-12T15:12:43Z

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

@@ -0,0 +1,1250 @@
+{


PDBFixer should be able to deal with Path objects but it does not now. We should open an issue for that.
Also, can we parameterize the pH value with an optional keyword? (..., pH=7.0):

Reply via ReviewNB

jaimergp · 2021-02-12T15:12:43Z

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

@@ -0,0 +1,1250 @@
+{


Capitalize Python.

Reply via ReviewNB

jaimergp · 2021-02-12T15:12:43Z

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

@@ -0,0 +1,1250 @@
+{


I thought OpenForceField could convert to OpenMM just fine. Am I missing something?

Reply via ReviewNB

Since we have to convert the unit of the positions we extract topology and positions separately and create a modeller object from there. Maybe a modeller object could be created from the openff Molecule directly, but what about the unit conversion then? We could probably do it at another step of the process, if that's better. Any ideas?
OpenFF Molecule Docs

jaimergp · 2021-02-12T15:12:43Z

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

@@ -0,0 +1,1250 @@
+{


Are you sure you need that +1 in complex_positions[len(protein.positions) + 1: total_atoms]? Python indexing is [start, end), so I would assume that we can start the 2nd range where the 1st ended, not shifted +1. What's the extra atom? You can also drop 0 and total_atoms. In other words:

complex_positions = unit.Quantity(np.zeros([total_atoms, 3]), unit=unit.nanometers) complex_positions[:len(protein.positions)] = protein.positions # add protein positions complex_positions[len(protein.positions):] = ligand.positions # add ligand positions

Reply via ReviewNB

jaimergp · 2021-02-12T15:12:43Z

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

@@ -0,0 +1,1250 @@
+{


Link to TIP3P and GAFF references, maybe?

Capitalize Python.

Reply via ReviewNB

Didn't find anything suitable about tip3p, especially not in context with amber. Added a link to the wikipedia article about water models which includes a section for tip3p. Better recommendations are welcome

I think this is the canonical citation: https://aip.scitation.org/doi/10.1063/1.445869

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

jaimergp · 2021-02-12T15:12:43Z

teachopencadd/talktorials/019_md_simulation/talktorial.ipynb

@@ -0,0 +1,1250 @@
+{


This # NBVAL_CHECK_OUTPUT will almost certainly fail because the minimization is a stochastic process and we are drawing velocities from a distribution, so for sure we will not get the same exact numbers every time.

Remove this comment and maybe, in a new cell, check if everything ran correctly by checking that the simulation objects has reached the desired number of steps.

Reply via ReviewNB

@Mika-Le

How about removing the #NBVAL_CHECK_OUTPUT from the simulation step cell and adding another cell afterwards with the following content:

simulation_time = round(simulation.context.getState().getTime().value_in_unit(unit.femtoseconds))
if simulation_time == steps * 2:
print("Simulation finished successfully!")
else:
print("Simulation failed!")
# NBVAL_CHECK_OUTPUT

jaimergp · 2021-02-12T15:14:04Z

This is awesome! I am sure OpenMM users will like this material a lot! Thanks for putting it together!

I have added some minor comments that do not change the outcome or purpose of the talktorial, just some technical details we need to take into account.

jaimergp · 2021-02-12T15:17:56Z

Ping me again when the comments are addressed so I resolve the conflicts and merge!

jaimergp · 2021-04-27T08:20:49Z

Thanks @Mika-Le! 🚀

Mika-Le added 2 commits October 8, 2020 12:43

Start branch

e63465f

Start Branch

d9cd2fa

jaimergp changed the base branch from master to packaging October 26, 2020 16:07

Mika-Le added 2 commits November 3, 2020 09:16

Code wrapped in functions

7575d21

Ready except loading the pdb.

b326df6

schallerdavid reviewed Nov 16, 2020

View reviewed changes

Mika-Le added 2 commits November 22, 2020 08:12

Theory ready for review

859d899

most code recommendations included

2374fa9

Base automatically changed from packaging to master November 23, 2020 10:42

Mika-Le and others added 3 commits November 25, 2020 09:05

theory on cadd added

3b2ecee

ready for review

0ceeb2a

polishing talktorial

d1ded86

schallerdavid requested a review from AndreaVolkamer November 26, 2020 21:26

schallerdavid added 4 commits December 2, 2020 14:40

switching to 5ug9

cfb3d5b

added to html and minor typos

922cf5d

switching to 3poz

ed3b2b6

image links to commit hashes on github

3e471e4

dominiquesydow added the new talktorial New talktorial label Dec 11, 2020

jaimergp marked this pull request as ready for review December 11, 2020 09:28

dominiquesydow changed the title ~~Ml 018 review~~ T018 - Molecular dynamics simulation Dec 11, 2020

dominiquesydow mentioned this pull request Dec 11, 2020

Base branch for new talktorials #74

Merged

27 tasks

dominiquesydow changed the title ~~T018 - Molecular dynamics simulation~~ T019 - Molecular dynamics simulation Dec 11, 2020

AndreaVolkamer reviewed Dec 17, 2020

View reviewed changes

Ready for second review

5384e6f

Ready for second review

27520a7

very minor textual adaptions

abd44a6

Presentation comments implemented. Ready!

7ab481f

schallerdavid added 2 commits February 12, 2021 13:45

minor changes and improved behavior for CI and Colab

1bd7cb0

add links to run talktorial on colab

5b51daf

jaimergp reviewed Feb 12, 2021

View reviewed changes

Mika-Le and others added 4 commits February 18, 2021 16:28

Most of Jaimes comments addressed. Check simulation output pending

16eb0f9

Merge branch 't011-base' into ml-018-review

382ff5d

rename 019 -> T019

8cede41

small style fixes

62ea4cb

jaimergp changed the base branch from master to t011-base April 27, 2021 08:19

jaimergp merged commit a5f38e2 into t011-base Apr 27, 2021

jaimergp deleted the ml-018-review branch April 27, 2021 08:20


		## Categories

		This talktorial is part of the following categories: [Collections overview](link)

T019 - Molecular dynamics simulation #56

T019 - Molecular dynamics simulation #56

Conversation

Mika-Le commented Oct 26, 2020 • edited Loading

Details

Content review

Code review

review-notebook-app bot commented Oct 26, 2020

schallerdavid commented Nov 26, 2020

jaimergp commented Nov 27, 2020

schallerdavid commented Dec 3, 2020

dominiquesydow commented Dec 11, 2020 • edited Loading

AndreaVolkamer Dec 17, 2020 • edited Loading

Choose a reason for hiding this comment

AndreaVolkamer Dec 17, 2020 • edited Loading

Choose a reason for hiding this comment

AndreaVolkamer Dec 17, 2020 • edited Loading

Choose a reason for hiding this comment

AndreaVolkamer Dec 17, 2020 • edited Loading

Choose a reason for hiding this comment

AndreaVolkamer Dec 17, 2020 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

AndreaVolkamer Dec 17, 2020 • edited Loading

Choose a reason for hiding this comment

AndreaVolkamer Dec 17, 2020 • edited Loading

Choose a reason for hiding this comment

Mika-Le Jan 11, 2021 • edited Loading

Choose a reason for hiding this comment

AndreaVolkamer Dec 17, 2020 • edited Loading

Choose a reason for hiding this comment

AndreaVolkamer Dec 17, 2020 • edited Loading

Choose a reason for hiding this comment

Minimize energy and output the minimized system

Choose a reason for hiding this comment

Choose a reason for hiding this comment

AndreaVolkamer commented Dec 17, 2020 • edited Loading

Mika-Le commented Jan 20, 2021

AndreaVolkamer commented Jan 24, 2021

jaimergp commented Jan 25, 2021

jaimergp commented Feb 2, 2021

schallerdavid commented Feb 12, 2021

Choose a reason for hiding this comment

jaimergp Feb 12, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jaimergp Feb 12, 2021 • edited Loading

Choose a reason for hiding this comment

jaimergp Feb 12, 2021 • edited Loading

Choose a reason for hiding this comment

jaimergp Feb 12, 2021 • edited Loading

Choose a reason for hiding this comment

jaimergp Feb 12, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jaimergp Feb 12, 2021 • edited Loading

Choose a reason for hiding this comment

jaimergp Feb 12, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jaimergp commented Feb 12, 2021

jaimergp commented Feb 12, 2021

jaimergp commented Apr 27, 2021

Mika-Le commented Oct 26, 2020 •

edited

Loading

dominiquesydow commented Dec 11, 2020 •

edited

Loading

AndreaVolkamer Dec 17, 2020 •

edited

Loading

AndreaVolkamer Dec 17, 2020 •

edited

Loading

AndreaVolkamer Dec 17, 2020 •

edited

Loading

AndreaVolkamer Dec 17, 2020 •

edited

Loading

AndreaVolkamer Dec 17, 2020 •

edited

Loading

AndreaVolkamer Dec 17, 2020 •

edited

Loading

AndreaVolkamer Dec 17, 2020 •

edited

Loading

Mika-Le Jan 11, 2021 •

edited

Loading

AndreaVolkamer Dec 17, 2020 •

edited

Loading

AndreaVolkamer Dec 17, 2020 •

edited

Loading

AndreaVolkamer commented Dec 17, 2020 •

edited

Loading

jaimergp Feb 12, 2021 •

edited

Loading

jaimergp Feb 12, 2021 •

edited

Loading

jaimergp Feb 12, 2021 •

edited

Loading

jaimergp Feb 12, 2021 •

edited

Loading

jaimergp Feb 12, 2021 •

edited

Loading

jaimergp Feb 12, 2021 •

edited

Loading

jaimergp Feb 12, 2021 •

edited

Loading