Pyter: Bayesian quantitative virology with Python and Numpyro

Pyter is an open-source Python package for analyzing virological data. It allows for flexible Bayesian inference of virus titers and virus environmental halflives from raw data on cell infection. Both endpoint titration and plaque assays are supported.

Pyter is designed to be fairly plug-and-play for the scientist who has a relatively standard problem but is a novice coder, but the API also permits a more experienced user to build their own custom models without reinventing the wheel.

PyTer uses Numpyro to specify models and perform inference from them.

Installing Pyter

Pyter is written in Python, and so to install it you will need a working Python 3 installation along with the standard package manager pip

Linux, macOS, and other Unix-like systems

You can use pip to install Pyter directly from the project GitHub::

pip install git+https://github.com/dylanhmorris/pyter.git

Alternatively, you can download the repository from GitHub, navigate to the top-level directory (containing pyproject.toml, and run:

pip install .