Chem 2920: Special Topics: Introduction to Machine Learning for Chemistry

We are excited to introduce a new Special Topics course focused on Machine Learning (ML) for Chemistry in Spring 2026

Who can take the course?

The course is designed for undergraduates in Chemistry and related majors. 

The course does not assume any prior experience in programming, machine learning, or Python.

The prerequisites are: Math 2374 (Multivariable Calculus), Chem 1072 (General Chemistry II), or equivalent course

The course will count as an Adv Tech Elective for Chemistry Majors

When is the course being offered?

The course will be offered in Spring 2026

Who should I contact for further information?

The instructor for the course is Dr. Sapna Sarupria (sarupria@umn.edu)

Know your instructor: Dr. Sarupria has been a faculty in the Chemistry department since 2021. She did her undergraduate in chemical engineering in India, and then moved to the US for graduate studies. She did her Master's in Texas A&M University (TX), PhD in Rensselaer Polytechnic Institute (NY), Postdoc in Princeton University (NJ), and started her academic career as a faculty in Clemson University (SC). Dr. Sarupria's research combines molecular simulations, machine learning, and advanced computational techniques to study molecular behavior of materials -- such as enzymes, viruses, polymers, and clays. In addition to research, Dr. Sarupria is passionate about inclusivity and bringing the human factor in STEM. Learn more about Dr. Sarupria and her research here. Her favorite things to do in her downtime are hanging out with her two cats, gardening, and reading books. 

Why should I take this course?

“AI will not replace humans, but those who use AI will replace those who don’t.” – Ginni Rometty, Former CEO of IBM 

Machine learning has permeated every aspect of our life -- from simple text editing to doing advanced research and development. 

Learning Machine Learning is now just as important as knowing the periodic table for Chemists. Even if you don't use machine learning directly in your job, you will interface with people who do. Therefore, being data- and machine learning-literate is not just important but necessary to be competitive in this market.

Lastly, it is fun to explore and learn -- it is a field that is rich with opportunities, and we are just getting started with using & developing Machine Learning for Chemistry, Molecules, and Materials.  The soil is fertile! 

To provide further context, the 2024 Nobel Prizes in Physics AND Chemistry were given to Artificial Intelligence!

"In October 2024, the Nobel Committees in Stockholm announced that the prizes in Physics and Chemistry were awarded to work related to artificial intelligence (AI). The prize in Physics was awarded to John J. Hopfield and Geoffrey E. Hinton (formerly of Google) for “foundational discoveries and inventions that enable machine learning with artificial neural networks.” The prize in Chemistry was awarded one-half to David Baker for “computational protein design” and one-half to Demis Hassabis and John M. Jumper (of DeepMind) for “protein structure prediction”"

Quote from: "Artificial Intelligence awarded two Nobel Prizes for innovations that will shape the future of medicine", Ben Li & Stephen Gilbert, npj Digital Medicine volume 7, Article number: 336 (2024)

What will I learn in the course?

This course introduces undergraduate students to the fundamentals of machine learning (ML) with a focus on its applications in chemistry. The course is designed to introduce basic concepts and terminology of machine learning, illustrate basic approaches of machine learning as applied to a broad spectrum of problems relevant to chemistry, and demonstrate the strengths and limitations of machine learning for Chemistry. 

Course Goals and Objectives:

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

1. Describe the key concepts, terminology, and approaches relevant to machine learning in chemistry.

2. Write basic code in Python relevant to data analysis and machine learning. 

3. Develop basic data handling and analysis skills, as well as skills for data visualization in Python

4. Apply basic machine learning approaches to various problems relevant to chemistry.

5. Critically evaluate the choice and design of the ML methods based on the quality and quantity of available input data

What is the structure of the course and expected workload?

• The course includes short pre-assignments like reading articles and watching videos before the class. (1-3 hours per week).

• The in-class component is 3 hours a week (3-credit hour course). The classes will include short lectures, hands-on exercises. The exercises will be done on Jupyter Notebook using Google Colab (we will teach you how to use these tools).

• Homeworks are expected to require about 3-5 hours per week and will include reading literature, coding, and developing machine learning models pertinent to chemistry.

• In the last two weeks of the course, the focus will be on a team project. It is expected that you will put in 6-9 hours on the projects each week for two weeks. We will use the class time and office hours for team meetings with the instructor.

This course is divided into three modules:

Module 1: Introduction to Python and Basics of Mathematics for Machine Learning

This module will cover topics to introduce Python and relevant Python concepts required to do machine learning in Python. Additionally, the mathematical concepts relevant to machine learning, including linear algebra (vectors, matrices), calculus (derivatives, gradients), probability, and statistics, will be covered. The exercises relevant to the mathematical concepts will be performed in Python to provide both training synergistically. This module will also begin to give an introduction to data visualization and data analysis in Python.

Module 2: Introduction to Machine Learning with Simple Examples

This module will cover various topics related to machine learning. Topics will include basic terminology of machine learning, data curation, data cleaning, and data handling, feature engineering, supervised and unsupervised learning, training-testing-validation, dimensionality reduction, deep neural networks, and approaches to chemistry such as AlphaFold, large language models for Chemistry, etc.

Module 3: Project-Based Learning of Advanced Applications of Machine Learning to Chemistry

This module will be primarily project-based. The students will work in teams to apply machine-learning techniques to a problem of their choice. Examples could include, the assessment of various techniques for polymer property predictions, investigating the effects of features on predictions, exploring the architecture and effectiveness of methods like Alphafold etc. The projects will be determined in consultation with the instructor, and weekly meetings between the teams and the instructor will be arranged for regular check-ins.