SPIMquant is a Snakebids app for quantitative analysis of SPIM (lightsheet) brain data. It performs automated nonlinear template registration and quantification of pathology from SPIM microscopy datasets.

Features

• Deformable registration to a template
• Atlas-based quantification of pathology
• Coarse-grained and fine-grained parallelization using Snakemake and Dask
• Support for reading BIDS datasets directly from cloud-based object storage
• Support for simple and scalable cloud-based processing with Coiled

Hardware Requirements

Processing lightsheet microscopy data is computationally-demanding, and you will need sufficient (and ideally fast and local) disk space. The more cores you have access to, the faster the code will run, but you will also need sufficient memory (e.g. 2-4 GB per core) as well.

Software Requirements

A Linux machine with Singularity or Apptainer installed, and a recent version of Python (>= 3.10). The workflow will download any containers it requires to run non-python dependencies (c3d, greedy, ANTS).

Installation

Install the package directly from GitHub:

pip install -e git+https://github.com/khanlab/spimquant#egg=spimquant

Or if you are going to make changes to the code, clone the repository then install it:

git clone https://github.com/khanlab/SPIMquant.git
pip install ./SPIMquant

Usage

SPIMquant is a BIDS App, so you need a BIDS dataset containing SPIM (or lightsheet microscopy) data to use it. The SPIMprep workflow is the recommended tool to produce a BIDS dataset from your raw or minimally-preprocessed microscopy data.

Perform a dry run:

spimquant /path/to/bids/dir /path/to/output/dir participant -np --use-conda

Run the app using all cores:

spimquant /path/to/bids/dir /path/to/output/dir participant --cores all --use-conda

If your input BIDS dataset stores data in zarr zipstores (e.g. SPIM files ending in *_SPIM.ome.zarr.zip), then you should use the following option:

--filter-spim extension='ome.zarr.zip'

Full Documentation: here