Empowering Chemists with AI and ML: Transformative Tools and Techniques

DAY 1: Getting Started with AI/ML for Chemists

Foundations (Days 1–2)

DAY 1: Getting Started with AI/ML for Chemists

Lecture 1: Introduction to AI/ML in Chemistry

• What is AI/ML? Why does it matter for chemists?
• When to apply AI/ML in research & industry
• Importance of data and quality
• Data sources: PubChem, Reaxys, proprietary datasets
• Examples of successful AI/ML in chemistry

Lecture 2: Tools & Workflow Basics (Hands-on)

• Python essentials for chemists (numpy, pandas, matplotlib)
• Introduction to Jupyter & Deepnote (step-by-step setup & walkthrough)
• Loading, cleaning, and visualizing chemical datasets
• First ML demo: simple regression/classification on a chemistry dataset

DAY 2: Core ML Concepts for Chemistry

Lecture 3: Machine Learning Algorithms (Practical overview)

• Supervised vs unsupervised learning
• Classification vs regression (chemistry examples)
• Random Forest demo (predict molecular property from descriptors)
• PCA & clustering demo

Lecture 4: Practical Hands-On Workshop

• Step-by-step exercise in Deepnote:
  • Load dataset
  • Engineer descriptors (RDKit basics)
  • Train/test split, build a model
  • Evaluate results (visualizations, metrics)
•  Short industry reflection: where these methods help day-to-day

Homework / Self-Study Week : Slack/Teams/email channel open for Q&A (optional, but helps engagement).

DAY 3: Feature Engineering & Chemistry Integration

Lecture 5: Feature Engineering in Chemistry

• What is feature engineering, why it matters
• Descriptor design: numerical vs categorical
• Hands-on with RDKit: compute common descriptors
• Create training/test sets and evaluate performance

Lecture 6: Case Study 1 (Guided Workshop)

• chemistry dataset (participants follow line-by-line)
• Build pipeline: descriptors → ML model → interpret results
• Interactive Q&A

DAY 4: Case Studies & Advanced Applications

Lecture 7: Case Study 2 (Hands-On)

• Lipophilicity (logP values)
• Train ML model (classification/regression)
• Interpreting results in chemistry terms

Lecture 8: Where Next? Emerging Methods & Wrap-Up

• Integration with Lab Automation & Cloud Platforms
  • Using AI models alongside ELNs (e.g., Chemotion)
  • Lab automation + AI = self-driving experiments
  • Industry applications in formulation, process chemistry, drug discovery
• Explainable AI (XAI) for Chemistry – Beyond prediction: why a model makes a decision
• AutoML for Chemists: Platforms that automate ML pipelines

During the lectures, participants will be guided step by step on how to prepare the environment and load the datasets. For those who wish to set things up in advance, please refer to the Software section of the course notes (provided before the course) for detailed instructions on creating a Deepnote account and accessing the example notebooks.

• Deepnote (cloud-based Jupyter environment) – free account required, runs in any modern browser, no installation needed.
• Python packages (pre-installed in Deepnote): NumPy, pandas, scikit-learn, matplotlib, RDKit.
• Datasets: Free, open datasets

• Understand the fundamentals of AI/ML and its tailored application to chemistry.
• Explore the diverse applications of AI/ML in different branches of chemistry, such as drug discovery, molecular design, reaction prediction, and spectroscopy analysis.
• Learn about commonly used AI/ML algorithms and tools in chemistry research.
• Implement AI/ML techniques to solve real-world chemistry problems.
• Develop critical thinking skills to evaluate the potential impact of AI/ML on the future of chemistry research and industry.

• Graduate students and researchers in chemistry and related fields seeking to enhance their knowledge and skills in AI/ML.
• Professionals working in pharmaceuticals, materials science, chemical engineering, and other industries interested in leveraging AI/ML for innovation.
• Educators and instructors looking to integrate AI/ML concepts into their chemistry curricula.
• Policymakers and regulators involved in shaping the future of AI/ML applications in chemistry.