ECIFP - Extended Connectivity Interaction Fingerprints

An interaction fingerprint for analyzing protein-ligand contacts, binding site and binding mode similarities.

Overview

This project provides tools for:

• Generating Extended Connectivity Interaction Fingerprints (ECIFP) for protein-ligand complexes in the PDB
• Identifying and analyzing binding sites
• Computing binding mode and binding site similarity
• Comparing protein-ligand interaction patterns

Requirements

• Python ≥ 3.12
• uv package manager

Installation

1. Install uv (if not already installed)

# macOS/Linux
curl -LsSf https://astral.sh/uv/install.sh | sh

# Windows
powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

# With pip
pip install uv

2. Clone the repository

git clone https://github.com/PDBeurope/ecifp.git
cd ecifp

3. Create virtual environment and install dependencies

# Create a virtual environment and install all dependencies
uv sync

# This will:
# - Create a .venv directory
# - Install Python 3.12 if needed
# - Install all dependencies from pyproject.toml
# - Lock versions in uv.lock

4. Activate the virtual environment

# macOS/Linux
source .venv/bin/activate

# Windows
.venv\Scripts\activate

Running the Notebooks

Using Jupyter Lab (Recommended)

# Start Jupyter Lab
jupyter lab

# Navigate to the nbs/ directory and open any notebook

Notebooks

The nbs/ directory contains the following Jupyter notebooks:

Note

Before running the notebooks, download data from 10.5281/zenodo.18557398 and set the path to data in conf.ini file

1. generate_ecifp.ipynb

Generates Extended Connectivity Interaction Fingerprints (ECIFP) for protein-ligand complexes using their interaction data

2. generate_binding_site_fp.ipynb

Generates fingerprints for binding sites of proteins based on their ligand interacting residues

3. binding_site_identification.ipynb

Clusters ligands to distinct non-overlapping binding sites based on the similarity of their binding site fingerprints

4. binding_site_similarity.ipynb

Evaluating the performance of ECIFP for binding site comparisons using benchmark data from

5. binding_mode_similarity.ipynb

Demonstrating the use of ECIFP for binding mode simialrities of ligands

Project Structure

ecifp/
├── README.md                # This file
├── pyproject.toml           # Project dependencies and configuration
├── uv.lock                  # Locked dependency versions
├── ecifp/                   # Main Python package
│   ├── __init__.py
│   ├── fp_sim.py           # Fingerprint generation and similarity calculations
│   └── utils.py            # Utility functions
├── nbs/                    # Jupyter notebooks
│   ├── generate_ecifp.ipynb
│   ├── generate_binding_site_fp.ipynb
│   ├── binding_site_identification.ipynb
│   ├── binding_site_similarity.ipynb
│   └── binding_mode_similarity.ipynb
└── images/                  # figures

License

CC0 1.0 Universal

Citation

[Add citation information if applicable]