Setup and Installation

Dependencies

E3FP is compatible with Python 3.x. It additionally has the following dependencies:

Required

• NumPy

• SciPy

• RDKit

• mmh3

• python_utilities

• smart_open

Optional

The following packages are required for the specified features:

• parallelization:

  • mpi4py

• molecular standardisation:

  • standardiser

• protonation states:

  • cxcalc

• storing conformer energies:

  • h5py

• faster fingerprint metric calculations:

  • numba

Installation

The following installation approaches are listed in order of recommendation.

Option 1: Install with Pip

Basic installation:

$ pip install e3fp

With optional dependencies:

$ pip install e3fp[optional]

Option 2: Install from conda-forge

E3FP is available on conda-forge.

$ conda create -n e3fp_env -c conda-forge e3fp
$ conda activate e3fp_env

To install optional dependencies:

$ conda install -c conda-forge mpi4py h5py standardiser

Option 3: Install from source

1. Clone the repository:

   $ git clone https://github.com/keiserlab/e3fp.git
   $ cd e3fp
   

2. Install for development in an already-activated environment.

   You can do this using pip:

   $ pip install -e .[dev]
   

Or use uv to set up a development environment:

$ uv sync --extra dev

Testing

Run tests using pytest:

$ pip install pytest  # if not already installed
$ pytest e3fp