Plinder Data Loader

[tjduigna]

This PR grew to far too many things. Here is a summary of what's contained:

• The PlinderDataset pytorch loader that iterates plinder systems and emits structures
  • An optional [loader] block to install the additional dependencies (torch and atom3d)
  • This could be extended to emit ligand structures independently
• Pipeline hacks to get straggling requirements over the line
• A docker / CI bug fix introduced by the presence of the docs branch precluding from publishing on merges to main
• Bug fix to publicly available docker container that resolves a failure to import plinder
• Enhancements to the plinder.core API which includes many necessary changes for usability
• A download_plinder_cmd to download the entire dataset all at once
• Improved support for translating pyarrow-derived query filters to duckdb-SQL queries
• A full fledged PlinderSystem class for managing the files within the plinder dataset
• Usage of cloudpathlib for abstracting the remote / local file pathing in plinder
• Small bug fixes in the end-to-end pipeline for:
  • annotation generation and saving
  • crystal contact annotation and usage
  • binding affinity downloads
  • config serialization / deserialization
  • better names and more consistency for configuration parameters
  • saving linked structures
  • more consistent config usage across the library
  • linked structure evaluation
• Many more test fixtures and assets
• More tests of plinder.core
• An "end-to-end" test for the pipeline that captures up to protein score generation

[github-actions[bot]]

Coverage report

Click to see where and how coverage changed

File	Statements	Missing	Coverage	Coverage (new stmts)	Lines missing
src/plinder/core/index
utils.py					81-82
src/plinder/core/loader
loader.py					3-92
src/plinder/core/scores
ligand.py					66
protein.py					76-77
query.py					49, 73, 97, 104, 107, 176-178, 214, 216
src/plinder/core/split
utils.py					35-42
src/plinder/core/system
system.py					31, 56-57, 68, 138-141, 145-152, 157-160
utils.py					7-9, 19-39
src/plinder/core/utils
cpl.py					60-66, 71-76, 93-96, 105, 138-141, 153
src/plinder/data
_version.py
databases.py
get_system_annotations.py
save_linked_structures.py					14, 85-92, 133, 144-172, 181-196, 207-221, 359
splits.py					256
src/plinder/data/pipeline
io.py					141-149, 461, 466, 470
pipeline.py					53-56, 247, 259-260, 264, 268, 286
tasks.py					159, 285-286, 721, 728-729, 735-736, 754-771, 792-808, 1013-1030
utils.py					289, 482-485, 501-504, 508, 518
src/plinder/data/utils
tanimoto.py
src/plinder/data/utils/annotations
aggregate_annotations.py
extras.py
get_similarity_scores.py					366-373, 383-390, 396-403, 412-417, 422-424, 462-469, 479-494
interaction_utils.py
ligand_utils.py					1247
rdkit_utils.py
Project Total

The report is truncated to 25 files out of 42. To see the full report, please visit the workflow summary page.

This report was generated by python-coverage-comment-action