pyGRN

Artificial Gene Regulatory Network model and evolution in Python

Installation

Recommended installation method uses python's venv module:

git clone https://github.com/d9w/pyGRN.git
cd pyGRN
python -m venv [VENV_PATH]
source [VENV_PATH]/bin/activate
pip install -e .

where VENV_PATH is determined by the user (for example, ~/.venvs/pygrn)

pyGRN requires a grns directory and a logs directory. By default, these will be in the same directory that scripts are run from; for example:

cd pyGRN
mkdir grns
mkdir logs

GRN files saved in JSON format will be generated during evolution and placed in grns, while evolutionary logs will be written in logs.

Usage

Examples of usage can be found in regression.py, which can run various regression problems, and dqn.py, which uses keras-rl to perform deep Q learning with a GRN.

Testing

To run the tests, use pytest or pytest --cov=./ to get a coverage report. Certain tests are stochastic and should pass most of the time, but might fail simply due to random initialization (eg tests/grns/test_grn_compare.py). To run these tests with verbose output, use:

pytest -v -s tests/grns/test_grn_compare.py

The tests also have many examples of usage beyond the above files.