clonedetective

A python package for automated cell lineage analysis.

What does it do?

clonedetective is a package for analysing fluorescent imaging data from cell lineage experiments (e.g. FLP-out, MARCM or Cre-lox clones).

Outputted quantifications include:

• counts of each cell type per “clone”
• spatial metrics e.g. number of nearest neighbors of each cell per clone
• cell and clone properties e.g. area, mean intensity etc.

If clones label genetic mutations, these metrics can be useful in addressing biological questions such as:

• does my gene of interest regulate clone size (i.e. cell proliferation) or clone composition (i.e. cell differentiation)?
• does the local cell neighbourhood (e.g. number and type of neighbours) non-autonomously impact cell proliferation or differentiation?
• does my gene of interest regulate the expression of other (immunostained) proteins?

Under the hood

clonedetective is constructed using many amazing python libraries, including scikit-image, numpy, Xarray, pandas, numba, Dask, Dask-image and pyclesperanto-prototype.

Many functions are lazy-loaded and parallelized using Dask, enabling clonedetective to scale to large multi-dimensional datasets that do not fit in RAM.

Install

It is recommended to install clonedetective into a virtual environment e.g. using conda.

If you have anaconda or miniconda installed, you can create a virtual environment using the following command. It is often helpful to install something into an empty environment, in this case we install scipy:

conda create -n myenv scipy

We next activate our environment:

conda activate myenv

You can then install clonedetective into this environment via pip:

pip install clonedetective

How to use

Please see the tutorials:

1. Example walkthrough
2. Downstream Analysis using Python
3. Downstream Analysis using R
4. Generating nuclei segmentation using StarDist

In progress, more to come!