cenfind 0.9.6

Score cells for centrioles in IF data

cenfind

A command line interface to score cells for centrioles.

Introduction

cenfind is a command line interface to detect and assign centrioles in immunofluorescence images of human cells. Specifically, it orchestrates:

• the z-max projection of the raw files;
• the detection of centrioles;
• the detection of the nuclei;
• the assignment of the centrioles to the nearest nucleus.

Installation

1. Install python via pyenv
2. Download and set up 3.9.5 as local version
3. Install poetry, system-wide with pip install poetry

Check that you're at the correct location (a simple and recommended location is cd ~, i.e., your home folder).

4. Download cenfind with:

git clone [e-mail]:UPGON/cenfind.git
git clone [e-mail]:maweigert/spotipy.git

5. As of now, you need to install the spotipy package from the git repository https://github.com/maweigert/spotipy: !!! You need to have access to this private repo; contact Leo for setting up the permission.

cd cenfind

6. Activate the virtual environment using poetry

poetry shell

Your prompt should now be prepended with (cenfind-py3.9).

Note: if your python version is not supported, install the one recommended with pyenv, the set it up and run poetry env use $(which python). Then, repeat the step.

6. Add the programs squash and score to the PATH with the following commands, so that they can be run from the command line, without the need to type the whole path.

poetry install

6. Add manually the package spotipy

pip install -e ../spotipy/

7. Check that cenfind's programs are correctly installed by running:

squash --help

8. In case of updates, get the last version:

git pull
poetry install

Basic usage

Before scoring the cells, you need to prepare the dataset folder. cenfind assumes a fixed folder structure. In the following we will assume that the .ome.tif files are all immediately in raw/. Each field of view is a z-stack containing 4 channels (0, 1, 2, 3). The channel 0 contains the nuclei and the channels 1-3 contains centriolar markers.

<project_name>/
└── raw/

2. Run setup to initialise the folder with a list of fields and output folders:

prepare /path/to/dataset <list channels of centrioles, like 1 2 3, (0 should be the nucleus channel)>

2. Run squash with the argument of the path to the project folder and the suffix of the raw files. projections/ is populated with the max-projections *_max.tif files.

squash path/to/ds .ome.tif

3. Run score with the arguments source and the index of the nuclei channel (usually 0 or 3).

score /path/to/dataset ./model/master/ 0 1 2 3 --projection_suffix '_max'

4. Check that the predictions are satisfactory by looking at the folder outlines and at the results/scores.csv.

API

cenfind consists of two core classes: Dataset and Field.

A Dataset represents a collection of related fields, i.e., same pixel size, same channels, same cell type.

It should:

• return the name
• iterate over the fields,
• construct the file name for the projections and the z-stacks
• read the fields.txt
• write the fields.txt file
• set up the folders projections, predictions, visualisations and statistics
• set and get the splits

A Field represents a field of view and should:

• construct file names for projections, annotation
• get Dataset
• load the projection as np.ndarray
• load the channel as np.ndarray
• load annotation as np.ndarray
• load mask as np.ndarray

Using those two objects, cenfind should

• detect centrioles (data, model) => points,
• extract nuclei (data, model) => contours,
• assign centrioles to nuclei (contours, points) => pairs
• outline centrioles and nuclei (data, points) => image
• create composite vignettes (data) => composite_image
• flag partial nuclei (contours, tolerance) => contours
• compare predictions with annotation (points, points) => metrics_namespace