Add dgidb and pharmgkb for drug annotation

[wxicu]

PR Checklist

• [x] Referenced issue is linked #483
• [x] If you've fixed a bug or added code that should be tested, add tests!
• [ ] Documentation in docs is updated

Description of changes

• Add dgidb for drug annotation
• Fix bug and add new functions for lookup object with type drug
• The dgidb metadata is saved for enrichment function. A small example:

  adata = sc.datasets.pbmc3k_processed()
  pt_drug = pt.md.Drug()
  pt_drug.annotate(adata, source="dgidb")
  pt_enricher = pt.tl.Enrichment()
  pt_enricher.score(
  adata, targets = pt_drug.dgidb_dict
  )
  # pt_drug.dgidb_dict saves the target and compunds in a unnested dictionary.

• However currently pt.tl.Enrichment.plot_dotplot doesnt accept targets in the unnested format (gives error if an unnested dict is given). I think it makes sense to go with nested dict to filter further categories. For this the user can play with the metadata

  interaction = pt_drug.dgidb_df
  interaction["interaction_type"] = interaction["interaction_type"].fillna("Unknown/Other")
  nested_dict = (
  interaction.groupby("interaction_type")
  .apply(
      lambda x: x.groupby("drug_claim_name")["gene_claim_name"].apply(list).to_dict()
  )
  .to_dict()
  )

  I didnt save it into Drug() because there are multiple possible categories to choose including approved, immunotherapy, anti_neoplastic, interaction_type.

Do you probably have any suggestions how to interact with enrichment funtion better?

Technical details

Additional context

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 45.61404% with 93 lines in your changes are missing coverage. Please review.

Project coverage is 62.27%. Comparing base (0f888e8) to head (36beb40).

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #575 +/- ## ========================================== - Coverage 62.66% 62.27% -0.40% ========================================== Files 43 43 Lines 5164 5296 +132 ========================================== + Hits 3236 3298 +62 - Misses 1928 1998 +70 ``` | [Files](https://app.codecov.io/gh/theislab/pertpy/pull/575?dropdown=coverage&src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=theislab) | Coverage Δ | | |---|---|---| | [pertpy/metadata/\_drug.py](https://app.codecov.io/gh/theislab/pertpy/pull/575?src=pr&el=tree&filepath=pertpy%2Fmetadata%2F_drug.py&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=theislab#diff-cGVydHB5L21ldGFkYXRhL19kcnVnLnB5) | `84.21% <83.51%> (-3.29%)` | :arrow_down: | | [pertpy/tools/\_enrichment.py](https://app.codecov.io/gh/theislab/pertpy/pull/575?src=pr&el=tree&filepath=pertpy%2Ftools%2F_enrichment.py&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=theislab#diff-cGVydHB5L3Rvb2xzL19lbnJpY2htZW50LnB5) | `47.27% <0.00%> (-5.08%)` | :arrow_down: | | [pertpy/metadata/\_look\_up.py](https://app.codecov.io/gh/theislab/pertpy/pull/575?src=pr&el=tree&filepath=pertpy%2Fmetadata%2F_look_up.py&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=theislab#diff-cGVydHB5L21ldGFkYXRhL19sb29rX3VwLnB5) | `8.71% <3.22%> (-1.74%)` | :arrow_down: |

[wxicu]

@Zethson

1. df() and dict() function is added. If this structure is ok then i will adapt to other metadata classes
2. add two new parameters in plotting function. I am not good at naming, but hope the structure is what we expect
3. if all above is fine, i will add pharmgkb database soon