Initial Standardization module import

[Allend95]

Description

Add the Standardization module to Opencadd formely developed under the name standardizer and mrstan for Allen Dumler's Bachelor thesis.

Final Tasks

• Read papers (28.-30. April)
• Save citations (28.-30. April)
• integrate citations into text (30. April)
• complete Background chapter (30. April -1. May)
• review of Background by Jaime?
  • fixing remarks
• review curation steps (Trust, but verify paper) (1. May)
• complete data and methods chapter (1.-2. May)
• review of Data and Methods by Jaime?
  • fixing remarks
• review and complete explanational text in notebook (2. May)
• add minor missing functions to notebook (2. May)
• final review of notebook by Jaime?
  • fixing remarks
• evaluate resulting functionality in Results and Discussion (3. May)
• complete Results and Discussion (4. -5. May)
• review of Results and Discussion by Jaime?
  • fixing remarks
• finish Conclusion (6. May)
• finish Outlook (6. May)
• review whole Thesis gramatically + spell check (7. May)
• final review of the text by Jaime?
  • fixing final remarks

Todos CW 10

• Goal: Finished guided notebook
  • further Evaluation of Substeps mentioned in paper
  • write helper functions to cover filtering steps
  • write explanations for every standardization step done
• Commented list of content (describing what will be covered in each chapter
• fill out sign up form for thesis (Implementation and evaluation of a computanional standardization pipeline for chemical compounds)

Todos CW 11

• Research a good suiting data set for testing of the pipline
• Setup overleaf doc
  • create chapters
  • base necessities

Todos CW 12

Code:

• Check if test data sets trigger any filtering functions
• add explanational texts
• Write actual code for:
  • data import
  • SMILEs to Mol Conversion
  • detection of inorganics + Creation of subsets
  • detection of mixtures + Creation of subsets

Todos CW 13

Code:

• Write actual code for:
  • filtering metals
  • Structural Conversion and Cleaning
    • Concept
    • --> Realize code from it
  • Removal of salts
    • Concept
    • --> Realize code from it
  • Normalization of Specific Chemotypes
    • Concept
    • --> Realize code from it
  • Removal of duplicates
    • Concept
    • --> Realize code from it

Todos CW 14

• fixing some issues with subfunctions in the notebook
  Writing
• Abstract (first draft)
• Introduction (first draft)
• Reading and listing Ressources for Introduction

Todos CW 15

Writing:

• Introduction
  • turn notes into text
• Data and Methods (draft)

Todos CW 16

Writing:

• Data and Methods
  • turn notes into text
• Results and Discussion (draft)
• Review Abstract
• Write Conclusion
• Outlook (draft)

Todos CW 17

• Review the explanational text of the notebook
• Add missing explanations
• Write explanational text on:
  • Normalization of Specific Chemotypes
  • Removal of duplicates

Non Technical To Dos:

• Get my registration form ready for the Thesis
• Organising my literature list
  • Read through important papers / literature
  • Take organized notes and evaluate in which chapters of the BT to use
  • organize citations
    • set up citavi
    • set up the right citation style
• Continue Writing the chapters of the thesis in the overleaf doc (Goal finished Thesis)
  • Abstract
  • Introduction / Problems
  • State of the Art
  • Data and Methods
  • Results and Discussion
  • Conclusion
  • Outlook

Pull Request To Do's:

Todos

• Evaluation of test data sets
• Check what changed in rdKit that causes old pytest's to fail
• Finish Notebook based on Fourches 2010 Paper

Notable points that this PR has either accomplished or will accomplish.

• initial import of the old code
• rename the imports inside the standardization modules
• rename the modules following the OpenCadd naming convention
• adjust tests
• add wrapper functions for following functions
  • Chem.SetAromaticity(mol)
  • Chem.SetHybridization(mol)
  • AllChem.ComputeGasteigerCharges(mol)
  • AllChem.Compute2DCoords(mol) (necessary to create 2D-Coordinates for molecules entered with a SMILES string, where this information is missing)

Status

• Ready to go

[jaimergp]

I noticed that init.py is not used to import modules anymore. How do I do that now or how are modules initialized now?

Can you point me to an example where this is happening? You still need __init__.py files afaik, unless you are talking about namespace (sub)packages.

[Allend95]

I noticed that init.py is not used to import modules anymore. How do I do that now or how are modules initialized now?

Can you point me to an example where this is happening? You still need __init__.py files afaik, unless you are talking about namespace (sub)packages.

Yeah, I thinks that's what I mean. What I did in my project, was to import all subpackages like this:

# Add imports here
from .standardization import *
from .assign_stereochemistry import *
from .convert_format import *
from .detect_inorganic import *

But in all other packages no one uses those import as far as I can see it. My problem now is, that when I want to load my modules, like for example to perform pytest , then it can't find my modules and I don't know how to configure it in any other way...

[jaimergp]

No, we don't want star * imports like that because they can pollute the namespace. For pytest, use the full import path, down to the module:

from opencadd.compounds.standardization import X

[jaimergp]

Ah, I see you are using __all__. Then star * imports are fine, as long as they stay within opencadd.compounds.standardization. Add an __init__.py there if needed.

[jaimergp]

These should be relative imports. If it's not working then there's something else going on, but for sure absolute paths should not be needed.

[jaimergp]

The absolute paths are only relevant in the pytest suite.

[jaimergp]

Hi @Allend95, I have cleaned the notebook a bit:

• Applied black-nb
• Removed the need to check for your current working directory -- we handle that through the _dh variable
• Replaced failedAt with a more appropriate lambda that ties the global scope to local scope through a keyword argument. It's more verbose but less prone to errors.