NSF center creates opportunities for computer-assisted discovery

As technology evolves, researchers are finding powerful ways to integrate computation, automation, and artificial intelligence into their work. At the forefront of this transformation are chemists and chemical engineers from Carnegie Mellon University, who are playing a key role in the National Science Foundation (NSF) Center for Computer Assisted Synthesis (C-CAS).

The multi-institutional initiative brings together experts in synthetic chemistry, computational chemistry, computer science, and related fields to accelerate new reaction discoveries and drug development through cutting-edge computational tools and collaborative research.

"The core of C-CAS is to take advantage of these modern algorithms and rethink organic chemistry with the promise to make it easier, faster, and more efficient," says Olexandr Isayev, Carl and Amy Jones Professor in Interdisciplinary Science. "There's this need to transform chemistry to take advantage of this revolution in computation and algorithms and robotics."

C-CAS spans 17 schools, including Carnegie Mellon, and allows researchers to freely collaborate with each other and industry partners on projects related to computational chemistry. Isayev and Gabe Gomes, assistant professor of chemical engineering and chemistry, joined Carnegie Mellon in part to participate in C-CAS.

Gomes said with C-CAS, chemists could significantly shorten the research process.

"The usual materials discovery and research and development cycle is about 10 years and about $10 million," Gomes says. "I want to bring development time down to one year and development costs to below $100,000. I think it's possible, and we're getting closer and closer as a community."

Isayev and Gomes' labs are working toward that goal. Gomes and other members of his lab have built an AI system driven by large language models (LLMs) that can work with automated science facilities to design, carry out, and analyze chemical reactions. These reactions can create thousands of novel compounds in a short amount of time.

This will speed up the process of making drugs faster, making them better, and making them cheaper.

Gabe Gomes, Assistant Professor, Chemical Engineering, Chemistry

"There is a case where we run over 16,000 reactions, and we get over one million compounds," Gomes says. "We're going from running four or 10 or 20 reactions over the course of a campaign to now scaling to tens of thousands or even higher. This will speed up the process of making drugs faster, making them better, and making them cheaper."

Robert MacKnight, a graduate student studying chemical engineering, is helping develop the AI tool by teaching it to gather and learn information from existing chemistry research online. He says that working with C-CAS has enhanced his work by allowing him to access programs and instruments that he can implement into the LLM.

"C-CAS has been instrumental through its emphasis on packaging research as 'tools' for researchers," MacKnight says. "This framework has been particularly valuable for my work because it allows me to greatly expand the capabilities of LLM systems — since tools are abundant and developed in a way that can be learned from by an LLM."

The Isayev lab is working with the Ukrainian company Enamine to develop machine learning tools to predict the outcomes of chemical reactions before they are performed.

"These reactions have been actively used in production to synthesize some of the building blocks for drug discovery," Isayev says. "Thanks to these technological developments, we could have fantastic progress in medicine."

Nick Gao, a chemistry graduate student in the Isayev lab, has collaborated on the project. He says that the newest versions of these machine learning tools, like AIMNet2, can suggest options for chemical reactions quickly and effectively.

"AIMNet2 can tell you which reactions will be most favorable from the starting point of the project," Gao says. "What if you have 100 molecules and you want to do a large-scale screening? AIMNet2 can do it within a minute."

C-CAS also provides mentorship to graduate students and postdoctoral fellows. Liliana Gallegos, a chemistry postdoctoral fellow who works with Gomes and Isayev, started working with C-CAS as a graduate student at Colorado State University. She says C-CAS has been instrumental to her journey as a chemist.

"Being part of these collaborations has brought me a lot of confidence," Gallegos says. "I've gotten to come at chemistry from different points of view. And with the collaborations, C-CAS provides a sense of community."

In May 2025, C-CAS held an annual meeting at Carnegie Mellon, bringing in collaborators working on organic synthesis from across the United States.

"The most important thing about an NSF center like this is that the results are more than the sum of their parts," Gomes says. "It really is a multiplicative output we have for such a team effort."

For media inquiries, please contact Lauren Smith at lsmith2@andrew.cmu.edu.