Custom base

[sonjaleo]

Description

Basis PR to fork from for additional programming. Will be merged to master after finishing.

Todos

• [x] folder for custom_kinfraglib_notebooks
• [x] rename notebooks to kinfraglib_notebooks
• [x] create "/custom_kinfraglib_notebooks/1_1_custom_filters_unwanted_substructures"
• [x] create "/custom_kinfraglib_notebooks/1_2_custom_filters_drug_likeness"
• [x] create "/custom_kinfraglib_notebooks/1_3_custom_filters_synthesizability"
• [x] implement @AndreaVolkamer's review comments for 1_1 to 1_3 notebooks https://github.com/sonjaleo/KinFragLib/pull/11
• [x] ~create "/custom_kinfraglib_notebooks/1_4_custom_filters_user_defined_fragment"] if time allows~
• [x] create "/custom_kinfraglib_notebooks/1_4_custom_filters_pairwise_retrosynthesizability" https://github.com/sonjaleo/KinFragLib/pull/9
• [x] create "/custom_kinfraglib_notebooks/2_1_custom_filters_pipeline" https://github.com/sonjaleo/KinFragLib/pull/12
• [x] create"/custom_kinfraglib/2_2_analysis https://github.com/sonjaleo/KinFragLib/pull/13
• [ ] ~create "/custom_kinfraglib_notebooks/2_1_custom_filters_analysis_enumerated_library~
• [ ] @dominiquesydow: Apply formatter/linter to original notebooks and package https://github.com/NonaR/KinFragLib/issues/4
• [ ] @dominiquesydow: Add back CI for original notebooks before merging into volkamerlab/kinfraglib https://github.com/NonaR/KinFragLib/issues/6

Questions

• [x] nb 1_2 add QED value column to fragment_library df?

Status

• [x] Ready to go

[dominiquesydow]

Hi @NonaR, I had a quick look at parts of your code and it looks really nice :)

If you are interested, we could have a call where we set up together some formatting (black) and linting (pylint or flake8) functionalities which help you to standardize code formatting and detect PEP8 violations automatically.

You could start playing around with

conda install black pylint
black -l 99 kinfraglib
pylint kinfraglib
# while you are in the local verison of this repo path: https://github.com/NonaR/KinFragLib

or you wait for the call to do this together.

[sonjaleo]

Hi @NonaR, I had a quick look at parts of your code and it looks really nice :)

If you are interested, we could have a call where we set up together some formatting (black) and linting (pylint or flake8) functionalities which help you to standardize code formatting and detect PEP8 violations automatically.

You could start playing around with

conda install black pylint
black -l 99 kinfraglib
pylint kinfraglib
# while you are in the local verison of this repo path: https://github.com/NonaR/KinFragLib

or you wait for the call to do this together.

Hi @dominiquesydow,

yes I would love to do this together :) If you want we can do this before the workshop I have today at 4pm

[dominiquesydow]

CI fails because it cannot find kinfraglib.filters because we currently install from kinfraglib from https://github.com/volkamerlab/kinfraglib/archive/master.tar.gz see here

Two options to fix this:

1. Update environment.yaml with current branch: https://github.com/nonaR/kinfraglib/archive/custom-base.tar.gz - but that would mean we need to do this for every new branch
2. For the time being add a pip install -e kinfraglib before running the CI tests

     - name: Run tests
       shell: bash -l {0}
       run: |
         pip install -e kinfraglib
         PYTEST_ARGS="--nbval-lax --current-env --nbval-cell-timeout=900"
         pytest $PYTEST_ARGS

[dominiquesydow]

Yaaaaay, passing CI!!!! :tada:

[dominiquesydow]

Open questions on CI fail:

• 2_4: Why does rdkit upgrade to 2020.09.1 not work? CI
• 4_1: Why JSONError for PATH_COMBINATORIAL_LIBRARY (combinatorial library downloaded from zenodo)? CI

[dominiquesydow]

@NonaR, the CI runs now on your notebooks (in notebooks/custom_kinfraglib) only; hope this will make your workflow faster and less error prone.

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[AndreaVolkamer]

@sonjaleo I reviewed the 1_3 nb. Looks very good, I mainly commented on adding some more info here and there.

Update:

• also ran it locally: environment contains everything to run nb 1.3
• nb 1.3 runs smoothly
• checked the utils functions, they look well, small todo
  • double check that the docs are up to date, and check some of the wordings in there
  • do we need to many subscripts (.py) or should be combine some of them (tbd)

Todo:

• Update main Readme
• Add filter Readmes
• Link to the readme's if necessary (e.g. for the data warrior stuff in nb 1.3)