Liver pdo

[alexandriai168]

Created samples, omics, drugs, and experiments python and shell scripts. This is related to issue #371

[alexandriai168]

Build runs without any errors and validated via link ml. Ready for review

[jjacobson95]

@alexandriai168 this is looking great overall!

I've confirmed that it builds and validates correctly. Files have the correct columns and reasonable values. Schema runs as expected. Transcriptomics file is confirmed to be in TPM scale.

There are just a few updates needed for this to be ready to merge, do you mind adding these?

• liverpdo_experiments.tsv: improve_sample_id column should be int not float.
• Lets update the model type in liverpdo_samples.csv to be more specific than organoid - this should either be patient derived organoid or xenograft derived organoid. Also, can you replace line 245 of the schema file coderdata.yaml ("organoid") with "patient derived organoid" to reflect this update.
• Add rm liverpdo_experiments_for_curve_fitting.tsv to the end of the drugs.sh file - this is to ensure that this file isn't accidentally captured in the final data we use wildcards to bring data across datasets.
• build_all.py should be modified in a similar way to build_dataset.py.
  • Please add liverpdo to the list of datasets to run in line 43.
  • Please add liverpdo to the dataset_map dictionary in the process_docker function (~line 103)
  • Please update lines 64, 333, 335 (synapse related checks) and 412 (general check) to include liverpdo.

And actually, while we are already making updates to build_dataset.py and build_all.py, the latest version(s) of docker compose is expected to be called in a slightly different way (issue #309) . This was already fixed once, but somewhere along the way it was accidentally reverted (probably with a merge). Do you mind adding this fix back in to this branch?

• Change compose_command = ['docker-compose', ... to compose_command = ['docker', 'compose', ... in both build_all.py and build_dataset.py.

[alexandriai168]

@jjacobson95 Okay! Finished these tasks! Thank you for reviewing this!