v2.3.0

BIOMERO v2.3.0

Minor release, with major new feature: introducing remote storage in-place import capabilities to the BIOMERO.analyzer by connecting its analysis results to the BIOMERO.importer.

What's Changed

• 🎯 Import Infrastructure: New SLURM Import Results functionality enables in-place import of analysis results via biomero.importer, keeping processed data with raw data on remote storage instead of uploading to OMERO's binary storage. Includes fail-fast validation checks, comprehensive status tracking, write validation, and improved error handling across the import pipeline.
• 📦 Dependencies: Pinned numpy < 2.0 and updated zarr constraints for improved compatibility and stability with scikit-image and related dependencies.
• 🔧 Enhanced Reliability: Better conversion job failure detection, improved database conflict handling for concurrent operations, and reduced logging verbosity for cleaner output.
• 🧹 Development Tools: Added no-cleanup debugging options to support troubleshooting during workflow failures.

Key Improvements

• Results can now be imported directly from remote storage via biomero.importer (optional)
  • Better backup strategy: analyzed data stays with raw data on the same storage
• Stricter dependency pinning to prevent zarr v3 breaking our conversion (part of patch v2.2.1)
• Enhanced job failure detection prevents silent errors in workflows (particularly conversion)

Optional: Enable In-Place Result Imports

By default, v2.3.0 continues using the standard OMERO API upload method (like all previous versions) and remains fully compatible with existing deployments. However, you can now optionally enable in-place imports of result images via the setup guide - perfect for deployments that want to keep analyzed data with raw data on the same remote storage!
This of course requires that you have the BIOMERO.importer deployed as well, for example by using the full NL-BIOMERO setup!

BIOMERO Changelog: v2.2.0...v2.3.0
Scripts Changelog: NL-BioImaging/biomero-scripts@v2.2.0...v2.3.0