Spatial Transcriptomics to Common Coordinate Framework Registration

Overview

sptx_ccf_registration is an image processing pipeline for registering spatial transcriptomics data to the common coordinate framework (CCF). This pipeline consists of three modules:

1. Segmentation of MERFISH point cloud data consisting of cell label and position into regions
2. Registration of MERFISH labels to CCF labels
3. Dashboard for analyzing the registration through metrics and image visualization

Installation

1. Clone this repository:

   git clone https://github.com/AllenInstitute/sptx_ccf_registration
   

2. Navigate to the directory:

   cd sptx_ccf_registration
   

3. Install the package:

   pip install .
   

4. Install dependencies for Neuroglancer

   pip install -r requirements_neuroglancer.txt
   

5. For Neuroglancer, obtain AWS credentials and export to your environment

Usage

For detailed usage instructions, see the documentation for individual modules in docs

You can use the following commands to get help on the input params for each module:

python -m sptx_ccf_registration.segmentation --help
python -m sptx_ccf_registration.registration --help
python -m sptx_ccf_registration.metrics_dashboard --help

You can configure these parameters with an input_json file for each module. Examples of these can be found in the sample_config directory.

Here are examples of launching the modules with the input_json files.

python -m sptx_ccf_registration.segmentation --input_json sample_config/segmentation/segmentation_broad.json
python -m sptx_ccf_registration.segmentation --input_json sample_config/segmentation/segmentation_landmark.json
python -m sptx_ccf_registration.registration --input_json sample_config/registration/registration.json
python -m sptx_ccf_registration.metrics_dashboard --input_json sample_config/metrics_dashboard/metrics_dashboard.json