FastTargetPred v2.0

Caution

⚠️ IMPORTANT DISCLAIMER - USE AT YOUR OWN RISK ⚠️

This repository has been ENTIRELY modernized and updated using AI assistance. The original FastTargetPred codebase has been significantly modified, including:

• Complete code refactoring and modernization
• New dependency management systems
• Updated Python syntax and type annotations
• New SMILES input functionality
• Integrated results analysis tools

While extensive testing has been performed, users should exercise extreme caution:

• ✅ Test thoroughly with your specific use cases before relying on results
• ✅ Validate outputs against known reference compounds
• ✅ Compare results with the original FastTargetPred version where possible
• ⚠️ Use at your own risk for research and production environments
• ⚠️ No warranty is provided for accuracy or functionality

The original research and algorithms probably maybe remain intact, but the implementation has been extensively modified.

FastTargetPred is a modern target prediction tool for molecular compounds that has been updated to use contemporary Python practices and dependency management.

Key Improvements in v2.0

1. Modern Dependency Management

• Added pyproject.toml for modern Python package configuration
• Added requirements.txt for backward compatibility
• Proper dependency specification with version constraints
• Support for optional dependencies (development, jupyter)

2. Updated Python Version Support

• Minimum Python version: 3.8+ (previously 3.7+)
• Modern type annotations throughout the codebase
• Updated syntax using f-strings and modern Python features
• Better error handling and type safety

3. Integrated MayaChemTools Support

• No longer requires MayaChemTools to be in PATH
• Automatically uses bundled mayachemtools directory
• Fallback to system PATH if bundled version not available
• Improved error messages and setup guidance

4. SMILES String Input Support

• New feature: Accept SMILES strings directly as input
• Format: name1:SMILES1 name2:SMILES2 or just SMILES
• Automatic conversion to SDF format using RDKit
• 3D coordinate generation and optimization
• Mutually exclusive with SDF file input

Installation

Prerequisites

1. MayaChemTools Setup: Download and unzip MayaChemTools into the mayachemtools/ folder:

   # The mayachemtools directory should be present in the root directory
   # If not already bundled, download from the official source and unzip here

2. Research Dataset (Optional): For access to the complete research dataset including 160,000 tetrapeptides and ChEMBL29 compounds:

   # Download the Zenodo research dataset
   wget https://zenodo.org/records/5751499/files/Zenodo_Final-data-_for_FastTargetPred-and-Tetrapeptides_data.zip
   unzip Zenodo_Final-data-_for_FastTargetPred-and-Tetrapeptides_data.zip

   The dataset contains:

   • 714,780 ChEMBL29 compounds with pre-computed fingerprints
   • 160,000 systematically enumerated tetrapeptides
   • Target prediction results and similarity analyses

Using pip (Recommended)

# Install in development mode
pip install -e .

# Or install from requirements.txt
pip install -r requirements.txt

Using conda (Alternative)

# Create a new environment
conda create -n fasttargetpred python=3.8
conda activate fasttargetpred

# Install dependencies
conda install numpy matplotlib pandas
conda install -c conda-forge rdkit

# Install the package
pip install -e .

Usage

Command Line Interface

Using SDF Files (Traditional)

# Basic usage with SDF file
fasttargetpred input.sdf

# With custom database and fingerprint
fasttargetpred input.sdf -db db/chembl25_active -fp ECFP4 MACCS

# Multiple SDF files
fasttargetpred file1.sdf file2.sdf file3.sdf

Using SMILES Strings (New Feature)

# SMILES with names
fasttargetpred -smiles "aspirin:CC(=O)OC1=CC=CC=C1C(=O)O" "caffeine:CN1C=NC2=C1C(=O)N(C(=O)N2C)C"

# SMILES without names (uses SMILES as name)
fasttargetpred -smiles "CC(=O)OC1=CC=CC=C1C(=O)O" "CN1C=NC2=C1C(=O)N(C(=O)N2C)C"

# Tetrapeptide example (from research dataset)
fasttargetpred -smiles "LAAA:[H]N[C@@H](CC(C)C)C(=O)N[C@@H](C)C(=O)N[C@@H](C)C(=O)N[C@@H](C)C(O)=O" -fp ECFP4 -tc 0.5

# Another tetrapeptide example - AEDG (Ala-Glu-Asp-Gly) using lower threshold for better coverage
fasttargetpred -smiles "AEDG:N[C@@H](C)C(=O)N[C@@H](CCC(O)=O)C(=O)N[C@@H](CC(O)=O)C(=O)NCC(O)=O" -fp ECFP4 -tc 0.3

Processing Results

# Generate HTML report from CSV output
fasttargetpred-workup -i output.csv -o report.html

# With structure images (requires RDKit)
fasttargetpred-workup -i output.csv -o report.html -sdf input.sdf

# With statistics plots
fasttargetpred-workup -i output.csv -o report.html -p

Programmatic Usage

from src.main import main
from src.arg_parsing import UserArguments

# Use the modernized API
# (Implementation details would depend on specific use case)

Dependencies

Required

• Python 3.8+
• NumPy >= 1.20.0
• Matplotlib >= 3.4.0
• RDKit >= 2022.3.1 (for SMILES support)
• Pandas >= 1.3.0

Optional

• Jupyter >= 1.0.0 (for notebook support)
• pytest >= 6.0 (for development)
• black >= 22.0 (for code formatting)
• flake8 >= 4.0 (for linting)
• mypy >= 0.910 (for type checking)

Key Changes from v1.x

Breaking Changes

• Minimum Python version increased from 3.7 to 3.8
• Some internal APIs have changed (mainly affects programmatic usage)
• Command-line interface remains largely compatible

New Features

• SMILES string input support with -smiles flag
• Automatic 3D coordinate generation for SMILES inputs
• Better error messages and validation
• Modern type annotations throughout

Improvements

• No longer requires MayaChemTools in PATH
• Better dependency management with pyproject.toml
• Cleaner, more maintainable code structure
• Enhanced error handling and user feedback

Migration Guide

From v1.x to v2.0

1. Update Python: Ensure you're using Python 3.8 or later
2. Install dependencies: Use pip install -r requirements.txt or pip install -e .
3. MayaChemTools: No longer needs to be in PATH - uses bundled version
4. SMILES support: New feature - try it with -smiles flag instead of SDF files

Command Line Compatibility

Most existing command lines should work without changes:

# v1.x (still works in v2.0)
python FastTargetPred.py input.sdf -db db/chembl25_active

# v2.0 (preferred)
fasttargetpred input.sdf -db db/chembl25_active

Development

Setting up Development Environment

# Clone and install in development mode
git clone https://github.com/ludovicchaput/FastTargetPred.git
cd FastTargetPred
pip install -e .[dev]

# Run tests
pytest

# Format code
black src/

# Type checking
mypy src/

Building C Extensions

# Compile platform-specific C extensions
python setup.py /path/to/python/include

License

This project maintains the same license as the original FastTargetPred.

Authors

• Valentin Guillaume
• Ludovic Chaput

Changelog

v2.0.0

• ✨ New: SMILES string input support
• ✨ New: Modern dependency management with pyproject.toml
• ✨ New: Integrated MayaChemTools (no PATH requirement)
• 🔧 Changed: Minimum Python version 3.8+
• 🔧 Changed: Modern type annotations throughout
• 🔧 Changed: Improved error handling and user feedback
• 🐛 Fixed: Various type safety and compatibility issues

v1.x

• Initial release with SDF file support
• Basic target prediction functionality

Original README

What does FastTargetPred do ? FastTargetPred allows to predict putative protein target(s) for a query compound or for a collection of compounds.

How to install FastTargetPred ?

1. Download the entire folder
2. Download the free chemistry toolkit MayaChemTools directory (http://www.mayachemtools.org/)
3. To launch FastTargetPred, you must use Python version 3.7 or superior. The help program includes every instruction to run the program. You can run it with the following command: python3.7 FastTargetPred.py -help

For more information, a user guide is provided: FastTargetPred_manual.pdf.