CoFD-Bench

Covalent docking benchmark system, containing structural data, analysis tools, and visualization scripts.

Project Overview

CoFD-Bench is a benchmark platform for evaluating covalent docking methods, containing information about ligands, warheads, Apo, Holo structures for multiple systems, as well as RMSD data for all docking results. This project provides a series of analysis tools for evaluating and comparing the performance of different docking methods.

Data Description

• pdb_benchmark.csv: Contains SuCOS-related data and distance information
• pdb_benchmark_RMSD.csv: Contains RMSD data for various docking methods
• plifs_recovery_df.csv: Contains PLIF (Protein-Ligand Interaction Fingerprints) recovery rate data

Analysis Features

This project implements various analysis and visualization functions:

• analyze_rmsd_statistics(): Comprehensive RMSD analysis with boxplots for different docking methods
• plot_combined_af3_analysis(): Integrates three AF3 analysis plots into one large figure
• plot_SuCOS_pocket_all(): Multi-method success rate analysis in 2x2 grid layout
• plot_combined_plif_analysis(): Combined bar and violin plots for PLIF recovery
• plot_sucos_group_PLIF(): Grouped stacked bar plots comparing docking outcomes by SuCOS groups

Example Jupyter Notebooks

• RMSD_plot.ipynb: RMSD-related analysis examples
• SuCOS_pocket_qcov.ipynb: SuCOS pocket coverage analysis
• PLIF_plot.ipynb: PLIF recovery rate analysis examples

Environment Setup

Create the project environment using conda:

conda env create -f environment.yml
conda activate CoFD-Bench

Usage Instructions

1. Activate the environment

conda activate CoFD-Bench

2. Run analysis scripts or Jupyter notebooks

jupyter notebook RMSD_plot.ipynb

3. Use functions from plot.py for custom analysis

import plot

# Basic RMSD analysis
plot.analyze_rmsd_statistics(
    csv_file='pdb_benchmark_RMSD.csv',
    output_dir='./results'
)

Output Formats

All analysis functions generate high-quality figures in multiple formats:

• PNG: High-resolution raster images (600 DPI)
• PDF: Vector graphics for publications
• SVG: Scalable vector graphics for web use