Welcome to CHEM 5080: AI for Experimental Chemistry

Washington University in St. Louis
Fall 2025
Instructor: Zhiling “Zach” Zheng (he/him)
Office: McMillen 411 | Web: Deep Synthesis Lab | Email: z.z@wustl.edu

This class is open to everyone and can be taken at your own pace.

• All tutorials are free and shareable.

• Created for synthetic chemists with no background in coding or data science.

• Focuses on practical applications of AI in wet lab synthesis and experimental data analysis.

• Lecture slides are available to registered students on Canvas or upon request.

If you find this course helpful, please consider sharing it with your friends.

Feedback and error reports are welcome by email.

Cite as:
Z. Zheng. Developing an AI Course for Synthetic Chemistry Students. arXiv:2511.18244, 2025.

---

Contents

• Lecture 1 - Python Primer
• Lecture 2 - Pandas and Plotting
• Lecture 3 - SMILES and RDKit
• Lecture 4 - Chemical Structure Identifier
• Lecture 5 - Regression and Classification
• Lecture 6 - Cross-Validation
• Lecture 7 - Decision Trees and Random Forests
• Lecture 8 - Neural Networks
• Lecture 9 - Graph Neural Networks
• Lecture 10 - Property & Reaction Prediction
• Lecture 11 - Dimension Reduction for Data Visualization
• Lecture 12 - Self-Supervised Learning
• Lecture 13 - De Novo Molecule Generation
• Lecture 14 - Bayesian Optimization for Synthesis Conditions
• Lecture 15 - Multi-Objective Bayesian Optimization
• Lecture 16 - Reinforcement Learning
• Lecture 17 - PU Learning
• Lecture 18 - Contrastive Learning
• Lecture 19 - Transformers
• Lecture 20 - Large Language Models
• Lecture 21 - Computer Vision
• Lecture 22 - Vision Languge Models
• Lecture 23 - Prompt Engineering and Function Calling
• Lecture 24 - Literature Data Mining
• Lecture 25 - Self-Driving Labs

---

Weekly Schedule

• Aug 26 & 28 – Intro + Python basics &

• Sept 2 & 4 – Molecular representations &

• Sept 9 & 11 – Classification & regression &

• Sept 16 & 18 – Supervised machine learning &

• Sept 23 & 25 – Molecular property & reaction prediction &

• Sept 30 & Oct 2 – Unsupervised learning methods &

• Oct 7 - Fall Break

• Oct 9 – De novo molecule generation

• Oct 14 & Oct 16 - Chemical reaction optimization &

• Oct 23 – Reinforcement learning

• Oct 28 & 30 – Semi-supervised learning &

• Nov 4 & 6 – Transformers & large language models &

• Nov 11 & 13 – Computer vision and multimodal models &

• Nov 18 & 20 – Multi-agent AI and literature data mining &

• Nov 25 – Self-driving labs

• Nov 27 – Thanksgiving

• Dec 2 & 4 – Final project presentations

Assignments & Project Submissions (Canvas)

Date	Assignment / Project
Sept 7	Assignment 1 due
Sept 21	Assignment 2 due
Oct 5	Project 1 due
Oct 14	Project 1 Evaluation due
Oct 26	Assignment 3 due
Nov 9	Assignment 4 due
Nov 30	Assignment 5 due
Dec 7	Project 2 Evaluations due

Course Materials

• All resources provided via this Jupyter Book

• Exercises in Google Colab notebooks

• Recommended (not required) readings:

  • Introduction to Python Computations in Science and Engineering (Kitchin)

  • Machine Learning in Chemistry (Janet & Kulik)

  • Deep Learning for Molecules and Materials (White)

Learning Outcomes

By the end of this course, you will be able to:

• Select and apply ML techniques for chemical problems.

• Visualize and interpret chemical data.

• Implement code for reaction optimization.

• Use computer vision for chemical data.

• Experiment with generative and transformer-based models.

Grading

• Homework (40%): 5 assignments with coding, data analysis, and visualization tasks.

• Midterm Project (20%): Mini-review essay (ACS style).

• Final Project (30%): Team presentation and executive summary.

• Peer Review (10%): Structured feedback on classmates’ work.

• Bonus Points (up to 5%): In-class polls, questions, and discussions.

Academic Policies

• Integrity: Cite all sources (including AI tools).

• Recording: No unauthorized recording or distribution of course materials.

• Accommodations: Students needing accommodations should contact Disability Resources.

• Equity & Support: University policies on harassment, reporting, and academic resources apply.

s

Acknowledgements

We thank Professor Robert Wexler for his guidance in building this website and for helpful discussions on the course content.
We also acknowledge Bingcui Guo, whose contributions were tremendously valuable in discussing the material and in testing the code.
This course was inspired by the insightful mind of Omar Yaghi, and we hope it encourages synthetic chemists to learn and apply AI in their research.