"Machine Learning for Drug Discovery" - Code and Data Repository

👋 Welcome to the Machine Learning for Drug Discovery Repository

This repository contains code and data for the first edition of Machine Learning for Drug Discovery (Manning Publications). The companion material within this repository covers introductory topics at the intersection of machine learning, deep learning, and drug discovery applied to real world scenarios in each chapter. The code and notebooks are released under the Apache 2.0 license.

For readability, the chapter notebooks only contain runnable code blocks and section titles. They omit the rest of the material in the book, i.e., text paragraphs, figures (unless generated as part of one of the code blocks), equations, and pseudocode. If you want to be able to follow what's going on, I recommend reading the notebooks side-by-side with your copy of the book!

Encounter any issues? Please let me know -- I can't fix a problem if I am not aware of its existence!

📚 Table of Contents

💊 Part 1: Fundamentals of Cheminformatics & Machine Learning

• Chapter 1: The Drug Discovery Process
• Chapter 2: Ligand-based Screening: Filtering & Similarity Searching
• Chapter 3: Ligand-based Screening: Machine Learning
• Chapter 4: Solubility Deep Dive with Linear Models
• Chapter 5: Classification: Cytochrome P450 Inhibition
• Chapter 6: Case Study: Small Molecule Binding to an RNA Target
• Chapter 7: Unsupervised Learning: Repurposing Drugs, Curating Compounds, & Screening Fragments

🧬 Part 2: Deep Learning for Molecules & Structural Biology

• Chapter 8: Introduction to Deep Learning
• Chapter 9: Structure-based Drug Design with Active Learning
• Chapter 10: Generative Models for De Novo Design
• Chapter 11: Graph Neural Networks for Drug Target Affinity Prediction
• Chapter 12: Transformer Architectures for Protein Structure Prediction
• Chapter 13: Multimodal AI Systems for End-to-End Drug Discovery Pipelines

Appendices

• Appendix A: Glossary
• Appendix B: Chemical Data Repositories
• Appendix C: Knowledge Distillation: Shrinking Models for Efficient, Hierarchical Molecular Generation
• Appendix D: Technical Deep Dive into Protein Structure Prediction

💻 Getting Started

Option 1: Google Colab (No Installation Required)

Open any notebook in Colab and run the installation cells at the top!

Each notebook includes two Colab installation options:

• Quick Install: Fast pip-based setup (3-10 minutes) with only the packages needed for that chapter
• Full Install: Complete conda environment (15-20 minutes) with all packages for all chapters

Option 2: Local Installation

Prerequisites: Python 3.12+ and git

We provide tiered installation options so you can install only what you need:

Core Environment (Chapters 1-4) — Basic ML & QSAR

git clone https://github.com/nrflynn2/ml-drug-discovery.git
cd ml-drug-discovery
pip install -r requirements-core.txt

Includes: numpy, pandas, matplotlib, seaborn, rdkit, scikit-learn

Advanced Environment (Chapters 5-8) — Gradient Boosting & Deep Learning

pip install -r requirements-advanced.txt

Adds: torch, xgboost, lightgbm, catboost, shap, umap, statsmodels

Full Environment (Chapters 9-11) — Molecular Docking & GNNs

conda env create -f ml4dd2025.yml
conda activate ml4dd2025

Adds: openmm, vina, pdbfixer, torch-geometric, mdtraj, prolif, meeko

Note: Chapters 9-11 require conda due to specialized packages (molecular dynamics, docking) that don't install reliably via pip.

Quick Reference:

• Chapter 1-4: Use requirements-core.txt
• Chapter 5-8, Appendix C: Use requirements-advanced.txt
• Chapter 9-11: Use ml4dd2025.yml (conda required)
• Chaper 12: Follow instructions and use the notebooks within CH12_FLYNN_ML4DD
• All chapters: Use ml4dd2025.yml for complete setup

For detailed installation instructions and troubleshooting, see INSTALL.md.

👥 Contribution & Support

Feel free to contribute, raise issues, or propose enhancements to make this repository a comprehensive resource for everyone venturing into machine learning, drug discovery, and related applications.

🔎 Citations

If you wish to cite the book, you may use the following:

@book{flynn2025mldd,
title={Machine Learning for Drug Discovery},
author={Flynn, N.},
isbn={9781633437661},
url={https://www.manning.com/books/machine-learning-for-drug-discovery},
year={2025},
publisher={Manning Publications}
}