1.23.0

Added

• h5ad cluster data extraction to IngestPipeline (#276)

1.22.0

Added

• Direct integration of ExpressionWriter into IngestPipeline (#273, #275)

1.21.0

Added

• Scratch scripts for generating data assets for image pipeline (#269, #270)

Fixed

• Correctly filter previously inserted DataArray documents during automatic reconnect retries (#271)

1.20.1

Changed

• Update basic AnnData validation to accept .h5 and .hdf5 file extensions (#266)

1.20.0

Added

• Basic AnnData ingest, confirm Scanpy compatibility (#264)

Changed

• DE output file naming to differentiate between results that only differ by '+' or '-' (#265)

1.19.2

Fixed

• fix imprecise sanitization on DE filename generation (#260)
• fix NA handling in DE analysis for correct filename generation (#261)

1.19.1

Fixed

• Fix ingest pipeline user_log.txt delocalization (#258)

1.19.0

Added

• Added duplicate gene handling to DE analysis (#256)

1.18.0

Added

• Delocalize DE analysis files to study bucket (#254)

Fixed

• Enable DE runs by filtering out annotation labels with only a single observation (#253)

1.17.2

Changed

• Updated DE to indicate annotation source, to handle DE for cluster annotations (#251)
• Revert generat try-except (#250)