Install & run

If you are starting from scratch, we recommend installing Anaconda.

MINT has been tested with Python 3.10, 3.11 and 3.12.

Then, download the latest release and extract it into a dedicated folder, or clone the repo with :

$ git clone https://github.com/biophotlumin/mint.git
$ cd mint

Then create a dedicated environment with :

$ conda create --name mint python=3.10
$ conda activate mint

Then install MINT with :

$ pip install .

You can also install optional dependencies by passing one of the following as :

$ pip install .[optional]

notebook

Jupyter dependencies to use interactive notebooks.

bioformats

ImageJ interface to read Bio-Formats extensions.

gui

GUI dependencies.

solvers

CVXPY and associated solvers for experimental trajectory denoising.

all

All of the above.

If you have trouble installing cvxpy with pip, try using conda.

For development, you can install in editable mode :

$ pip install -e .[optional]

You can then install :

docs

Sphinx and other dependencies to build docs.

dev

All of the above, including docs.

CLI

You can run MINT with :

$ mint -f <input_folder> -p <path/to/file> -l -e -s

-f, --folder

Optional. Path to raw data folder.

-p, --params

Optional. Path to config file (YAML or JSON).

-l, --locate

Optional. Run tracking.

-e, --extract

Optional. Extract transport parameters from trajectories.

-s, --stats

Optional. Run statistical analysis.

Ideally, the input folder should be specified in the config file (see the user guide).

If the config file does not contain an input path and --folder isn’t specified, the script will default to the current working directory.

If only --folder is specified, the script will look for a config file in that folder. Otherwise, it will fall back to default parameters.

When running --extract without --locate first, the input folder must contain the results of a previous --locate run.

Similarly, when running --stats without --extract first, the input folder must contain the results of a previous --extract run.

If neither --locate, --extract or --stats are specified, the script will go through a full run.

For the sake of reproducibility, you can download the paper branch from the repo, then import the environment we used in Grimaud et al. 2022 from reproducible_env.yml.