Add PharmaFormer model

[nictru]

A Transformer-based deep learning model designed to predict clinical drug responses
by integrating gene expression profiles and drug molecular structures.

Original authors: Zhou et al. (2025, 10.1038/s41698-025-01082-6)
Code adapted from their Github: https://github.com/zhouyuru1205/PharmaFormer

---

I investigated their source code thoroughly (as it was one of the most promising models for my internship task) and found the following problems in their evaluation:

1. Uses LPO fold creation
2. Does not account for Simpson's paradox in evaluation

Within drevalpy these parts are handled by the library, allowing us to see the actual performance, which is pretty bad.

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

❌ Patch coverage is 90.80780% with 33 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (development@8ad96bc). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
drevalpy/models/PharmaFormer/pharmaformer.py	94.27%	11 Missing ⚠️
tests/conftest.py	63.33%	11 Missing ⚠️
.../featurizer/create_pharmaformer_drug_embeddings.py	88.57%	8 Missing ⚠️
tests/test_featurizers.py	86.95%	3 Missing ⚠️
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@              Coverage Diff               @@
##             development     #336   +/-   ##
==============================================
  Coverage               ?   82.23%           
==============================================
  Files                  ?       65           
  Lines                  ?     6795           
  Branches               ?        0           
==============================================
  Hits                   ?     5588           
  Misses                 ?     1207           
  Partials               ?        0           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[nictru]

Yes they seem to have made very questionable choices, and I followed them very closely

If you don't evaluate properly it is not necessary to have a meaningful model to achieve 'good' results and end up in Nature

[PascalIversen]

Ah, ok, wow!!! Well in that case, I am looking forward to seeing how it performs when properly evaluated :D

[nictru]

I don't get a 'Merge' button, so if someone could please click it for me 😄

[JudithBernett]

I don't get a 'Merge' button, so if someone could please click it for me 😄

After you've added it to the docs :D

docs is missing

[JudithBernett]

add to docs please :)

[JudithBernett]

lgtm!