Update the polypharmacy example and GripNet implementation

[NYXFLOWER]

This PR will update and test the polypharmacy gripnet example and the core API related to GripNet. A tutorial notebook for Gripnet will be added in the next PR after this one is merged.

Change summary:

• In the core API
  • remove the old GripNet implementation and test.
  • add PolypharmacyDataset to kale.loaddata.polypharmacy_datasets
  • add DistMultDecoder and GripNetLinkPrediction to kale.predict.decode
  • add negative_sampling and typed_negative_sampling to kale.prepdata.graph_negative_sampling
  • add out_channels to GripNet in kale.embed.gripnet
  • add auprc_auroc_ap to kale.evaluate.metrics
• update the example and configs
• update README and docs

Follow-up future works in order of priority

• Add an example dataset with fixed training, validation, and test sets and update this example.
  • The variable names should be made more descriptive than the current version (g_, gg_, d_, gd_). Also be more consistent whether gene or protein.
  • Fix: some side effects do not have pairs in the test set.
• Provide functionalities to split graph data into training, validation, and test sets.
• Add Jupyter Notebook example.
• Improve test coverage.

Status

Ready

Types of changes

• Non-breaking change (fix or new feature that would not break existing functionality).
• Breaking change (fix or new feature that would cause existing functionality to change).
• New tests added to cover the changes.
• In-line docstrings updated.
• Source for documentation at docs manually updated for new API.

[codecov-commenter]

Codecov Report

Merging #346 (35ebc9b) into main (de78831) will decrease coverage by 1.42%.
The diff coverage is 38.75%.

@@            Coverage Diff             @@
##             main     #346      +/-   ##
==========================================
- Coverage   92.81%   91.38%   -1.43%     
==========================================
  Files          46       48       +2     
  Lines        5078     5130      +52     
==========================================
- Hits         4713     4688      -25     
- Misses        365      442      +77     

Impacted Files	Coverage Δ
kale/loaddata/polypharmacy_datasets.py	0.00% <0.00%> (ø)
kale/prepdata/graph_negative_sampling.py	26.66% <26.66%> (ø)
kale/evaluate/metrics.py	71.42% <33.33%> (-28.58%)	⬇️
kale/predict/decode.py	63.44% <53.42%> (-36.56%)	⬇️
kale/embed/gripnet.py	86.62% <100.00%> (-3.38%)	⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[haipinglu]

Many thanks for the great efforts. Almost there, with a few minor issues remaining.

Shall we ask @JunseKim1159 to help with unit tests to improve the coverage?

[NYXFLOWER]

Many thanks for the great efforts. Almost there, with a few minor issues remaining.

Shall we ask @JunseKim1159 to help with unit tests to improve the coverage?

Why not. @JunseKim1159 will be very helpful if he is interested. Besides the code format and testing cases, he can probably help with reconstructing a new valid polypharmacy example dataset and writing the notebook!

[haipinglu]

We'd better talk through the remaining problems to move forward. Many thanks.

[haipinglu]

I thought we have corrected the above variable named. g_, gg_, d_, gd_. Is it else where? Can we make them more descriptive here? Also gene or protein, again. I saw you used protein more.

[haipinglu]

This is due to the variable name in the dataset, to be updated in another PR.