locpix 0.1.0.3

Package for analysing SMLM data

locpix

locpix is a Python library for analysing point cloud data from SMLM.

This project is under active development

This is a short ReadMe just containing a QuickStart guide. For more comprehensive documentation please see https://oubino.github.io/locpix/

locpix includes the following functionality in order they are used in a normal workflow:

1. Preprocess : Initialises project and converts .csv files representing SMLM data (point cloud) into .parquet files, necessary for this software

2. Annotate : Generating histograms from the SMLM data and manually annotating these histograms to extract relevant localisations & labelling with seeds for watershed algorithm

3. Segmentation:

   1. Classic segmentation : Use classic method to segment histograms to extract relevant localisations

   2. Cellpose segmentation : Use Cellpose method to segment histograms to extract relevant localisations

   3. Ilastik segmentation : Use Ilastik method to segment histograms to extract relevant localisations

   4. UNET segmentation : Use UNET to segment histograms to extract relevant localisations

4. Membrane performance : Performance metrics calculation based on the localisations (not the histograms!)

Project Structure

We assume your input SMLM data are .csv files.

This input data must first be preprocessed into a user chosen project directory, using the Preprocess script. We strongly suggest this project directory is located outside the locpix folder.

The input and output of all further scripts will remain located inside the project directory, the input data folder will not be accessed again!

Usage configuration

Each script needs a configuration file (.yaml file), which should be specified using the -c flag.

Each configuration used will be saved in the project directory.

The templates for the configuration files can be found in the templates folder.

For instructions for the parameters in the configuration files see https://oubino.github.io/locpix/user_guide/templates/index.html

Quickstart

Installation

Prerequisites

You will need anaconda or miniconda or mamba. We recommend mamba

Install

Create an environment and install via pypi

conda create -n locpix-env python==3.10
conda activate locpix-env
pip install locpix

Preprocessing

Preprocess

This script preprocesses the input .csv data for later use AND must be run first.

This script will take in .csv files, and convert them to .parquet files, while also wrangling the data into our data format.

To run the script -i -c and -o flags should be specified

preprocess -i path/to/input/data -c path/to/config/file -o path/to/project/directory

Optionally we can add:

• -s flag to check the correct files are selected before preprocessing them

• -p flag if our files are .parquet files not .csv files.

Annotate

This script allows for manual segmentation of the localisations and adding markers to the images as seeds for the watershed algorithm.

To run the script -i and -c flags should be specified

annotate -i path/to/project/directory -c path/to/config/file

Optionally we can add:

• -m flag to check the metadata of the project before running

• -r flag to relabel histograms, will assume some labels are already present

Image segmentation

Classic segmentation

Perform classic segmentation on our localisation dataset.

To run the script -i and -c flags should be specified

classic -i path/to/project/directory -c path/to/config/file

Optionally we can add:

• -m flag to check the metadata of the project before running

Cellpose segmentation

Need to activate extra requirements - these are big and not included in initial install.

Note that if you have a GPU this will speed this up.

Note we modified Cellpose to fit in with our analysis, therefore you need to install our forked repository - note below will clone the Cellpose repository to wherever you are located

If you have a GPU

pip install torch torchvision --extra-index-url https://download.pytorch.org/whl/cu117
git clone https://github.com/oubino/cellpose
cd cellpose
pip install .

If you don’t have a GPU

pip install pytorch
git clone https://github.com/oubino/cellpose
cd cellpose
pip install .

To evaluate Cellpose model on the localisation dataset without any retraining on your dataset run the script with -i and -c flags specified

cellpose_eval -i path/to/project/directory -c path/to/config/file

Optionally we can add:

• -m flag to check the metadata of the project before running

• -o flag to specify folder in project dir to save output (defaults to cellpose_no_train)

• -u flag to specify a user model to load in

To retrain first then evaluate we instead

Prepare data for training

train_prep -i path/to/project/directory -c path/to/config/file

Optionally we can add:

• -m flag to check the metadata of the project before running

Train cellpose

cellpose_train -i path/to/project/directory -ct path/to/config/train_file -ce path/to/config/eval_file

Optionally we can add:

• -m flag to check the metadata of the project before running

UNET segmentation

Need to activate extra requirements - these are big and not included in initial install.

Note that if you have a GPU this will speed this up.

Note this is only needed if haven’t done for cellpose above

If you have a GPU

pip install torch torchvision --extra-index-url https://download.pytorch.org/whl/cu117

If you don’t have a GPU

pip install pytorch

To train UNET

unet -i path/to/project/directory -c path/to/config/file

Optionally we can add:

• -m flag to check the metadata of the project before running

Ilastik segmentation

Need to prepare the data for Ilastik segmentation

ilastik_prep -i path/to/project/directory -c path/to/config/file

Optionally we can add:

• -m flag to check the metadata of the project before running

Then run the data through the Ilastik GUI, which needs to be installed from Ilastik and to run it please see https://oubino.github.io/locpix/user_guide/usage.html#id7

Then convert the output of the Ilastik GUI back into our format

ilastik_output -i path/to/project/directory

Optionally we can add:

• -m flag to check the metadata of the project before running

Membrane performance

To evaluate membrane performance for a particular method, run below, where method name needs to match where the segmentation files are

membrane_performance_method -i path/to/project/directory -c path/to/config/file -o method_name

Optionally we can add:

• -m flag to check the metadata of the project before running

To evaluate performance of membrane segmentation from classic, cellpose and ilastik

membrane_performance -i path/to/project/directory -c path/to/config/file

Optionally we can add:

• -m flag to check the metadata of the project before running

To aggregate the performance over the folds for methods classic, cellpose without training, cellpose with training and ilastik

agg_metrics -i path/to/project/directory

Licenses

Package	License
alabaster 0.7.12	UNKNOWN
app-model 0.1.1	BSD 3-Clause License
appdirs 1.4.4	MIT
arrow 1.2.3	Apache 2.0
astroid 2.12.13	LGPL-2.1-or-later
asttokens 2.2.0	Apache 2.0
attrs 22.1.0	MIT
Babel 2.11.0	BSD
backcall 0.2.0	UNKNOWN
backports.functools-lru-cache 1.6.4	UNKNOWN
beautifulsoup4 4.11.1	MIT
binaryornot 0.4.4	BSD
black 22.12.0	MIT
build 0.9.0	MIT
cachey 0.2.1	BSD
cellpose 2.1.2.dev26+g731fe4e	BSD
certifi 2022.9.24	MPL-2.0
cfgv 3.3.1	MIT
chardet 5.1.0	LGPL
charset-normalizer 2.1.1	MIT
click 8.1.3	BSD-3-Clause
cloudpickle 2.2.0	BSD 3-Clause License
colorama 0.4.6	UNKNOWN
comm 0.1.3	BSD 3-Clause License
commonmark 0.9.1	BSD-3-Clause
contourpy 1.0.6	BSD-3-Clause
cookiecutter 2.1.1	BSD
coverage 6.5.0	Apache 2.0
cycler 0.11.0	BSD
dask 2022.11.1	BSD
debugpy 1.6.4	MIT
decorator 5.1.1	new BSD License
distlib 0.3.6	Python license
docstr-coverage 2.2.0	MIT
docstring-parser 0.15	MIT
docutils 0.17.1	public domain, Python, 2-Clause BSD, GPL 3 (see COPYING.txt)
entrypoints 0.4	UNKNOWN
exceptiongroup 1.0.4	UNKNOWN
executing 1.2.0	MIT
fastremap 1.13.3	LGPLv3
filelock 3.9.0	UNKNOWN
flake8 6.0.0	MIT
fonttools 4.38.0	MIT
freetype-py 2.3.0	UNKNOWN
fsspec 2022.11.0	BSD
HeapDict 1.0.1	BSD
hsluv 5.0.3	MIT
identify 2.5.17	MIT
idna 3.4	UNKNOWN
imagecodecs 2022.9.26	BSD
imageio 2.22.4	BSD-2-Clause
imageio-ffmpeg 0.4.7	BSD-2-Clause
imagesize 1.4.1	MIT
importlib-metadata 6.6.0	UNKNOWN
in-n-out 0.1.6	BSD 3-Clause License
iniconfig 1.1.1	MIT License
ipykernel 6.17.1	UNKNOWN
ipython 8.13.2	BSD-3-Clause
ipython-genutils 0.2.0	BSD
jedi 0.18.2	MIT
Jinja2 3.1.2	BSD-3-Clause
jinja2-time 0.2.0	MIT
joblib 1.2.0	BSD
jsonschema 4.17.3	MIT
jupyter-client 7.4.7	UNKNOWN
jupyter-core 5.1.0	UNKNOWN
kiwisolver 1.4.4	UNKNOWN
lazy-object-proxy 1.8.0	BSD-2-Clause
line-profiler 4.0.2	BSD
llvmlite 0.39.1	BSD
locket 1.0.0	BSD-2-Clause
locpix 0.0.12.dev70+ga7833b4.d20230120	UNKNOWN
magicgui 0.6.1	MIT license
MarkupSafe 2.1.1	BSD-3-Clause
matplotlib 3.6.2	PSF
matplotlib-inline 0.1.6	BSD 3-Clause
mccabe 0.7.0	Expat license
mypy-extensions 0.4.3	MIT License
napari 0.4.17	BSD 3-Clause
napari-console 0.0.6	BSD 3-Clause
napari-locpix 0.0.3	MIT
napari-plugin-engine 0.2.0	MIT
napari-svg 0.1.6	BSD-3
natsort 8.2.0	MIT
nest-asyncio 1.5.6	BSD
networkx 2.8.8	UNKNOWN
nodeenv 1.7.0	BSD
npe2 0.6.1	BSD-3-Clause
numba 0.56.4	BSD
numpy 1.23.5	BSD
numpydoc 1.5.0	BSD
opencv-python-headless 4.6.0.66	MIT
packaging 21.3	BSD-2-Clause or Apache-2.0
pandas 1.5.2	BSD-3-Clause
parso 0.8.3	MIT
partd 1.3.0	BSD
pathspec 0.10.2	MPL 2.0
pep517 0.13.0	UNKNOWN
pickleshare 0.7.5	MIT
Pillow 9.3.0	HPND
Pint 0.20.1	BSD
pip 23.1.2	MIT
platformdirs 2.5.4	UNKNOWN
pluggy 1.0.0	MIT
polars 0.15.1	MIT
pre-commit 3.0.3	MIT
prettytable 3.8.0	BSD (3 clause)
prompt-toolkit 3.0.33	UNKNOWN
psutil 5.9.4	BSD-3-Clause
psygnal 0.6.1	BSD 3-Clause License
pure-eval 0.2.2	MIT
pyarrow 10.0.1	Apache License, Version 2.0
pycodestyle 2.10.0	Expat license
pydantic 1.10.2	MIT
pydata-sphinx-theme 0.12.0	BSD 3-Clause License
pyflakes 3.0.1	MIT
Pygments 2.13.0	BSD License
PyOpenGL 3.1.6	BSD
pyparsing 3.0.9	UNKNOWN
PyQt5 5.15.7	GPL v3
PyQt5-Qt5 5.15.2	LGPL v3
PyQt5-sip 12.11.0	SIP
pyrsistent 0.19.2	MIT
pytest 7.2.0	MIT
pytest-cov 4.0.0	MIT
python-dateutil 2.8.2	Dual License
python-dotenv 0.21.0	BSD-3-Clause
python-slugify 7.0.0	MIT
pytomlpp 1.0.11	UNKNOWN
pytz 2022.6	MIT
PyWavelets 1.4.1	MIT
pywin32 304	PSF
PyYAML 6.0	MIT
pyzmq 24.0.1	LGPL+BSD
qtconsole 5.4.0	BSD
QtPy 2.3.0	MIT
requests 2.28.1	Apache 2.0
rich 12.6.0	MIT
scikit-image 0.19.3	Modified BSD
scikit-learn 1.1.3	new BSD
scipy 1.9.3	UNKNOWN
seaborn 0.12.2	UNKNOWN
setuptools 67.7.2	UNKNOWN
setuptools-scm 7.0.5	MIT
six 1.16.0	MIT
snowballstemmer 2.2.0	BSD-3-Clause
soupsieve 2.3.2.post1	UNKNOWN
Sphinx 4.5.0	BSD
sphinx-autoapi 2.0.0	MIT
sphinxcontrib-applehelp 1.0.2	BSD
sphinxcontrib-devhelp 1.0.2	BSD
sphinxcontrib-htmlhelp 2.0.0	BSD
sphinxcontrib-jsmath 1.0.1	BSD
sphinxcontrib-qthelp 1.0.3	BSD
sphinxcontrib-serializinghtml 1.1.5	BSD
stack-data 0.6.2	MIT
superqt 0.4.1	BSD 3-Clause License
text-unidecode 1.3	Artistic License
threadpoolctl 3.1.0	BSD-3-Clause
tifffile 2022.10.10	BSD
tomli 2.0.1	UNKNOWN
toolz 0.12.0	BSD
torch 1.13.0+cu117	BSD-3
torchsummary 1.5.1	UNKNOWN
torchvision 0.14.0+cu117	BSD
tornado 6.2	http://www.apache.org/licenses/LICENSE-2.0
tqdm 4.64.1	MPLv2.0, MIT Licences
traitlets 5.6.0	UNKNOWN
typer 0.7.0	UNKNOWN
typing-extensions 4.4.0	UNKNOWN
Unidecode 1.3.6	GPL
urllib3 1.26.13	MIT
virtualenv 20.17.1	MIT
vispy 0.11.0	(new) BSD
wcwidth 0.2.5	MIT
wheel 0.40.0	MIT License
wrapt 1.14.1	BSD
zipp 3.15.0	UNKNOWN