Add native Bliss synergy scoring

[daveringelberg]

LLM usage

• LLM or coding agent used: GPT-6.1 Sol Ultra via Codex
• If one was used, I, a human, have reviewed, understood and tested every changed line

PR Checklist

• Referenced issue is linked
• If you've fixed a bug or added code that should be tested, add tests!
• Documentation in docs is updated

Description of changes

Adds native Bliss scoring as a first combination-analysis step for #1036.

The proposed pt.tl.Synergy().score() takes AnnData with one observation per assay measurement for a fixed drug pair and experimental condition. Expected inhibition and observed-minus-expected scores are stored in .obs.

Technical details

Uses control-normalized inhibition fractions between 0 and 1. Single-agent replicates are averaged at exactly matching doses, without interpolation. Positive scores indicate excess inhibition, not statistical significance.

Additional context

Draft while the placement as a separate Synergy tool is discussed.

Validation: 39 relevant tests passed. Ruff, formatting, scoped type checking, executable examples and a focused Sphinx build passed.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 75.79%. Comparing base (a4c3073) to head (12d06ad).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1171      +/-   ##
==========================================
+ Coverage   75.69%   75.79%   +0.10%     
==========================================
  Files          55       56       +1     
  Lines        8435     8475      +40     
==========================================
+ Hits         6385     6424      +39     
- Misses       2050     2051       +1     

Files with missing lines	Coverage Δ
src/pertpy/tools/__init__.py	91.66% <100.00%> (+0.23%)	⬆️
src/pertpy/tools/_synergy.py	100.00% <100.00%> (ø)

... and 1 file with indirect coverage changes

🚀 New features to boost your workflow:

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