fractal-uzh-converters 0.5.1

A collection of fractal tasks to convert HCS Plates to OME-Zarr

Fractal UZH Converters

A collection of Fractal tasks to convert High-Content Screening (HCS) plate data from various microscopes into OME-Zarr format.

Supported Microscopes

Microscope	Manufacturer	Task Name
Opera Phenix / Operetta	PerkinElmer	Convert Operetta Plate to OME-Zarr
ScanR	Olympus	Convert Olympus ScanR Plate to OME-Zarr
CQ3K	Yokogawa	Convert Yokogawa CQ3K Plate to OME-Zarr
CellVoyager	Yokogawa	Convert Yokogawa CellVoyager Plate to OME-Zarr
ImageXpress HCS.ai	Molecular Devices	Convert MD ImageXpress HCS.ai Plate to OME-Zarr

Each converter reads the microscope's native metadata and image files, then produces a well-structured OME-Zarr HCS plate.

Installation

pip install fractal-uzh-converters

How It Works

Each converter is implemented as a Fractal Compound Task consisting of two steps:

1. Init task — Parses the microscope metadata, creates the OME-Zarr plate structure, and generates a parallelization list.
2. Compute task — Reads the raw image tiles and writes them into the OME-Zarr dataset. This task runs in parallel across wells.

You configure the init task with one or more acquisitions (paths to your raw data directories) and the converter handles the rest.

Key Features

• Multiple tiling modes — Snap to grid, snap to corners, inplace, or no tiling depending on your acquisition layout.
• Condition tables — Attach experimental metadata (drug treatments, concentrations, replicates) to wells via a CSV file.
• Flexible writer modes — Choose between per-FOV, per-tile, Dask-parallel, or in-memory writing strategies to balance speed and memory usage.
• Overwrite control — No overwrite, full overwrite, or extend mode to incrementally add acquisitions.
• OME-NGFF 0.4 and 0.5 — Target either specification version for the output.

Documentation

Full documentation is available at: https://fractal-analytics-platform.github.io/fractal-uzh-converters/