MotilA: A pipeline for microglial fine process motility analysis

Le résumé fourni par la source

🚀 MotilA v1.2.0 This release is a major internal refactor of the MotilA package, where we have modularized MotilA's monolithic codebase into focused submodules in order to improve maintainability and readability and to ensure future extensibility. We also exchanged the batch processing and collection functions with a more flexible subject-based discovery workflow that supports arbitrary folder-tag levels, image glob patterns, and exclusion filters. The new batch processor also supports skipping already processed outputs and continues processing after per-file errors. It also generates persistent root-level run reports and timestamped error reports for each batch run so that users can easily track the processing status of each file and subject. Since the batch_process_stacks and batch_collect functions have been redesigned, the previous pre-v1.2.0 implementations are still available as batch_process_stacks_old and batch_collect_old for backward compatibility, but they are now marked as deprecated. We recommend upgrading to MotilA v1.2.0 and updating batch scripts to the new processing_options and tag_folder_levels style. Internal package modularization MotilA's former monolithic motila.py implementation has been split into focused modules: motila.io for OMIO-backed image I/O. motila.export for Excel/CSV/YAML table exports. motila.projection for z-projections, projection plotting, and intensity output helpers. motila.preprocessing for stack inspection, subvolume extraction, registration, spectral unmixing, contrast normalization, and filtering. motila.segmentation for binarization, connected-component filtering, and pixel-area outputs. motila.motility for frame-to-frame motility metric calculation. motila.pipeline for the single-stack process_stack workflow. motila.batch for the current batch processing and collection functions. The historical motila.motila import path remains available as a compatibility and distributor layer. Existing scripts can continue importing and calling process_stack, batch_process_stacks, batch_collect, and lower-level helper functions from motila.motila. Development groundwork Preserved existing function signatures and docstrings during the split. Kept the current Zarr-backed intermediate processing unchanged. Added module-level docstring headers and # %% cell sections to match the project coding style. Verified the modularized package with the existing test suite and RTD build. Flexible BIDS-like batch processing The batch processor has been redesigned for flexible subject-based project trees. The new batch_process_stacks discovers image files below subject folders, supports arbitrary folder-tag levels, filters by image glob patterns, excludes preview/auxiliary files by name, skips already processed outputs when requested, and continues processing after per-file errors. New batch features: Flexible subject discovery via subject_ids or subject_prefix. Flexible tag_folder_levels, for example [("DC000_FOV", "DA000_FOV"), ("TL_000",)]. Default image patterns for TIFF/OME-TIFF, CZI, LSM and RAW files. exclude_name_contains filtering for files and matched tag folders. skip_processed=True to avoid reprocessing existing MotilA outputs. Persistent root-level run reports: motila_batch_run_report.yaml and motila_batch_run_report.txt. Timestamped root-level and per-file error reports. Structured result dataclasses with processed, skipped, failed and discovered records. Optional helper for creating Thorlabs RAW YAML sidecar templates from MotilA batch error reports. batch_collect now uses the same flexible discovery settings as batch_process_stacks, including subject selection, tag-folder levels, image patterns and exclusion filters. Upgrade note: The new flexible processor now uses the public name batch_process_stacks. The previous pre-v1.2.0 implementation is still available as batch_process_stacks_old and is marked as deprecated. The new flexible collector now uses the public name batch_collect. The previous pre-v1.2.0 collection implementation is still available as batch_collect_old and is marked as deprecated. Updating batch scripts to the new processing_options and tag_folder_levels style is recommended.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.