[Pilot] Generate Croissant metadata for pilot set of GEO-hosted datasets

[aclayton555]

Following the pilot in #208 , this effort will initiate investigation and generation of Croissant metadata for GEO-hosted datasets that are surfaced on the CCKP. This will be the first exploration of Croissant metadata generation for externally hosted datasets. This will be in parallel with expansion of Croissant metadata generation for Synapse-hosted datasets via #226 and in parallel with preliminary documentation efforts.

[aditya-nath-sage]

Created GEO dataset collection with 10 new datasets from multiple projects

Created a dataset in each project with the name of the GEO acession number. Then used this script to export sample level metadata from GEO: https://github.com/sagebio-ada/ADA_useful_scripts/blob/geo-metadata/data_processing/extract_geo_metadata.py

Then, uploaded those metadata files to synapse and added them as the single file in the dataset.

Once this is done, then I can add the newly created "dataset" entity to the dataset collection linked above.

Next steps: Need to link the correct schema to the datasets (Dataset Catalog schema that was refined by Orion and I for the croissant pilot), and then add metadata for each dataset. Once this is done, need to figure out how to parse the GEO metadata with the tech team to create the croissant metadata needed.

[aditya-nath-sage]

Update:

Attepted to create link entities for GEO linked sample data, but this data was only accessible via ENA or SRA. Re-creating link entities for one dataset to make sure that we have a working prototype for testing direct downloads from the CCKP
(https://www.synapse.org/Synapse:syn71549268.2/datasets/#query/eyJzcWwiOiJTRUxFQ1QgKiBGUk9NIHN5bjcxNTQ5MjY4LjIiLCAiaW5jbHVkZUVudGl0eUV0YWciOnRydWUsICJsaW1pdCI6MjV9)

[aditya-nath-sage]

Update:

Added a script which takes GEO sample accession #s and outputs links to SRA or ENA files: https://github.com/sagebio-ada/ADA_useful_scripts/pull/4/commits/a0860ba6f2cc06c17b4bcf0673b954390a95ec6b

Then went ahead and re-created link entities for the GEO dataset entity GSM277269: https://www.synapse.org/Synapse:syn71549268.3/datasets/
Had to create 2 link entities for each GSM # due to there being 2 files per accession # (paired read fastq files)