EASI-FISH Spots Tools

A Python toolkit for detecting and analyzing fluorescent spots in EASI-FISH imaging data. This package provides tools for spot detection using FishSpots, as well as utilities for extracting statistics from spots detected by various tools including airlocalize, RS-FISH, and FishSpots.

Installation

Using Conda

Create a conda environment with the required dependencies:

conda env create -f conda-env.yml
conda activate easifish-spots-tools

Install the package

pip install -e .

Requirements

• Python >= 3.12
• numpy > 2, < 3
• pandas = 2.3.3
• scikit-image = 0.25.2
• zarr
• dask/distributed
• See pyproject.toml for complete dependency list

Tools

1. Spot Detection

Run spot detection on image data using the FishSpots algorithm with distributed processing support.

Get help:

python -m easifish_spots_tools.main_spot_extraction -h

Basic usage:

python -m easifish_spots_tools.main_spot_extraction \
  --input /path/to/input.zarr \
  --output /path/to/output.csv \
  --channels 0 1 2 \
  --fishspots-config config.yaml \
  --local-dask-workers 4

Key parameters:

• --input: Path to input image container (Zarr/OME-Zarr format)
• --output: Path to output CSV file for detected spots
• --channels: Channel indices to process
• --fishspots-config: YAML configuration file for FishSpots parameters
• --local-dask-workers: Number of Dask workers for parallel processing
• --blocksize: Block size for processing (X,Y,Z)
• --intensity-threshold: Intensity threshold for spot detection
• --apply-voxel-spacing: Apply voxel spacing to spot coordinates

2. Extract Spot Counts

Count spots within labeled regions and aggregate results across multiple spot files.

Get help:

python -m easifish_spots_tools.labeled_spots_counts -h

Basic usage:

python -m easifish_spots_tools.labeled_spots_counts \
  --labels-container /path/to/labels.zarr \
  --labels-subpath /labels \
  --spots-pattern "/path/to/spots/*.csv" \
  --output counts.csv

Key parameters:

• --labels-container: Path to the labeled segmentation container
• --spots-pattern: Glob pattern for spot CSV files
• --timeindex: Time index for the labels
• --channel: Channel index for the labels
• --voxel-spacing: Voxel spacing (X,Y,Z) if not in metadata

3. Extract Spot Region Properties

Extract detailed region properties and statistics for spots within labeled regions.

Get help:

python -m easifish_spots_tools.labeled_spots_props -h

Basic usage:

python -m easifish_spots_tools.labeled_spots_props \
  --labels-container /path/to/labels.zarr \
  --image-container /path/to/image.zarr \
  --spots-pattern "/path/to/spots/*.csv" \
  --output properties.csv

4. Post-Processing for RS-FISH

Normalize spot data format across different detection tools (airlocalize, RS-FISH, FishSpots) to ensure consistent output format.

Get help:

python -m easifish_spots_tools.post_rs_fish -h

Purpose: This tool standardizes spot information regardless of the detection tool used, ensuring all spots follow the same CSV format with columns: x, y, z, t, c, intensity, sx, sy, sz.

Output Format

All tools produce CSV files with spot information in the following format:

• x, y, z: Spatial coordinates (in voxels or micrometers, depending on --apply-voxel-spacing)
• t: Time index
• c: Channel index
• intensity: Spot intensity
• sx, sy, sz: Spot sigma values in each dimension

Common Workflows

Full Pipeline: Detection to Analysis

1. Detect spots:

python -m easifish_spots_tools.main_spot_extraction \
  --input data.zarr \
  --output spots.csv \
  --channels 0 1 \
  --local-dask-workers 8 \
  --apply-voxel-spacing

2. Count spots per labeled region:

python -m easifish_spots_tools.labeled_spots_counts \
  --labels-container labels.zarr \
  --spots-pattern "spots.csv" \
  --output spot_counts.csv

3. Extract region properties:

python -m easifish_spots_tools.labeled_spots_props \
  --labels-container labels.zarr \
  --image-container data.zarr \
  --spots-pattern "spots.csv" \
  --output spot_properties.csv

Configuration

FishSpots Configuration

Create a YAML configuration file for spot detection parameters:

white_tophat_args:
  # White top-hat filter parameters

psf_estimation_args:
  # PSF estimation parameters

deconvolution_args:
  # Deconvolution parameters

spot_detection_args:
  # Spot detection parameters

Dask Configuration

For distributed processing, you can provide a Dask configuration file:

python -m easifish_spots_tools.main_spot_extraction \
  --dask-config dask_config.yaml \
  --dask-scheduler tcp://scheduler:8786

License

BSD-3-Clause

Version

0.3.0