AI-Enabled CADD & Machine Learning for Drug Design
From Molecular Representation, Property Prediction to Virtual Screening
Skills you will gain:
About Program:
Computer-Aided Drug Design has become an indispensable part of modern pharmaceutical research. By simulating molecular interactions, predicting ADMET properties, and screening millions of compounds virtually, CADD significantly speeds up early-stage drug discovery. With the addition of machine learning, researchers can now build data-driven models that enhance prediction accuracy, optimize hit selection, and streamline lead optimization.
This workshop provides a comprehensive introduction to structure-based and ligand-based drug design, molecular docking workflows, scoring functions, QSAR modeling, and ML algorithms used in cheminformatics. Participants will explore real datasets, learn to prepare protein/ligand structures, and perform computational experiments through hands-on dry-lab sessions.
Aim: This workshop aims to equip participants with a deep understanding of computational drug discovery workflows, integrating classical CADD methods with modern machine learning (ML) and AI-based predictive modeling. The goal is to provide hands-on experience with molecular docking, virtual screening, QSAR modeling, and ML-driven drug property prediction. The program prepares learners to utilize computational tools for accelerating drug discovery, reducing experimental costs, and generating high-confidence hit/lead candidates.
Program Objectives:
- Understand key concepts in structure-based and ligand-based drug design.
- Perform molecular docking, virtual screening, and scoring of ligands.
- Build QSAR and ML models for predicting biological activity & ADMET properties.
- Use cheminformatics tools for dataset preparation and feature extraction.
- Apply end-to-end computational workflows for early-stage drug discovery.
What you will learn?
Day 1 – CADD Basics & Molecular Representation
- Introduction to CADD in drug discovery: hit identification, optimization, ADMET.
- Ligand-based vs. structure-based design.
- Molecular representations: SMILES, SDF, descriptors, fingerprints.
- Basic CADD workflow: Data → Descriptors → Model/Docking → Evaluation.
- Tools: RDKit, Open Babel, PyMOL/ChimeraX (optional)
- Hands-on: Explore molecules and analyze properties (demo).
Day 2 – Machine Learning for Drug Design
- ML basics: features vs. labels, regression vs. classification.
- ML tasks in drug design: activity prediction, ADMET flags.
- Models: Linear/logistic regression, tree-based models.
- Evaluation: Accuracy, ROC-AUC, basic regression error.
- Tools: scikit-learn, XGBoost/LightGBM, DeepChem
- Hands-on: Design a small ML pipeline with descriptors and activity data.
Day 3 – Integrated CADD + ML Workflow
- Virtual screening & docking: scoring, ranking, and compound prioritization.
- Using ML to predict properties and prioritize compounds.
- Full workflow: Data → Descriptors → ML → Filter → Docking.
- Limitations: Data quality, overfitting, applicability domain.
- Tools: AutoDock Vina, PyRx, DeepChem, RDKit
- Hands-on:Design an integrated CADD+ML workflow for compound prioritization and docking.
Mentor Profile
Fee Plan
Get an e-Certificate of Participation!
Intended For :
- Undergraduate/postgraduate students in Biotechnology, Bioinformatics, Pharmacy, Chemistry, Computational Biology, or related fields.
- Researchers working in drug design, medicinal chemistry, cheminformatics, molecular biology, or wet-lab drug discovery.
- Professionals from pharma, biotech, CROs, and in-silico modeling platforms.
- Beginners with interest in computational drug discovery—no coding experience required.
Career Supporting Skills
Program Outcomes
- Gain practical knowledge of protein & ligand preparation for computational analysis.
- Learn to perform molecular docking, pose scoring, and affinity prediction.
- Build ML models for bioactivity and ADMET prediction.
- Work with real chemical datasets for QSAR modeling.
- Understand how CADD integrates with wet-lab drug discovery workflows.