Since the inception of the UChicago Pritzker School of Molecular Engineering, ongoing research has looked at the potential of artificial intelligence to develop artificial proteins, biocompatible electronics and other new technologies.
As AI technology has advanced, so have UChicago’s innovations.
In addition to using AI, UChicago PME researchers are also advancing the technology, charting new educational pathways for AI-assisted learning and, as UChicago PME Prof. Supratik Guha recently outlined for The Economic Times, building a roadmap for quantum computing to cut AI’s carbon impact.
UChicago PME Prof. Andrew Ferguson said AI is rapidly evolving to become a commodity tool in scientific inquiry.
“Nobody uses a slide rule anymore,” Ferguson said. “You use calculators or microcomputers. In the same way, AI is now becoming a standard tool for doing science.”
Here are a few ways AI at UChicago Pritzker Molecular Engineering, as part of the broader UChicago efforts in AI + Science, harnesses this new technology to tackle some of society’s biggest challenges.
Using computational modeling to tackle science’s biggest questions
Experts from around the globe gathered at the University of Chicago this summer to push the boundaries on using artificial intelligence, machine learning, and other computer modeling innovations to tackle the world’s biggest challenges.
The workshop Frontiers of Computational Reaction Prediction ran from July 15-17, bringing together researchers from different fields to promote fundamental research on advanced computational methods. It was the initial event hosted by CECAM-US-CENTRAL, the first North American node for the global research consortium Centre Européen de Calcul Atomique et Moléculaire (CECAM).
Ferguson, the director of the new node, said the time was right both for the event and for CECAM’s commitment to the North American research community.
“We’re at a point in history where high-performance computation, machine learning, and AI are powerful and pervasive tools for molecular and materials discovery that are being used for everything from finding candidates for sustainable battery materials to sorting through trillions of potentially lifesaving drugs and therapies,” Ferguson said.
Putting AI on the hunt for better batteries
As the globe transitions away from fossil fuels, more and better batteries will be needed to store the energy created by solar and wind power for times the sun isn’t shining and the wind isn’t blowing.
Bolstered by a new Google Research Scholar Award, UChicago PME Neubauer Family Assistant Professor Chibueze Amanchukwu is developing what he calls ElectrolyteGPT to sort through the millions upon millions of potentially useful chemical compounds to find better materials for battery electrolytes.
“We need clean energy solutions as soon as possible,” said Amanchukwu, whom the MIT Technology Review recently named an Innovator Under 35. “This effort could ultimately help us deliver solutions more quickly.”
A blueprint for training graduate students at the intersection of AI and materials science
Training leaders who can use AI to create a more sustainable future requires a new educational approach. A UChicago PME graduate training program called AI-enabled Molecular Engineering of Materials and Systems for Sustainability (AIMEMS) aims to do just that.
Together with similar programs from Duke University and the University of Illinois at Urbana-Champaign, AIMEMS faculty and staff wrote a Focus paper for the journal Science Advances, detailing their approach and successes.
“We found that we need to first provide coursework or boot camps to get everyone up to speed on AI and machine learning,” said UChicago PME Prof. Junhong Chen, a leader of the AIMEMS program who is also a lead water strategist at Argonne National Laboratory. “And then we need to facilitate interdisciplinary projects, but that is how students learn the most. Solutions across departments will always be better than solutions from one discipline.”
New research unites quantum engineering and artificial intelligence
The large carbon footprint and electricity cost of modern power-hungry machine learning and artificial intelligence models are motivating interests in rendering ML/AI cheaper and more sustainable.
An interdisciplinary team including PME Prof. Liang Jiang and CQE IBM postdoctoral scholar Junyu Liu, as well as researchers from UChicago CS, Argonne National Laboratory, UC Berkeley, MIT, Brandeis University and Freie Universität Berlin hope to change that.
In a paper published in Nature Communications, the team showed how incorporating quantum computing into the classical machine-learning process can potentially help make machine learning more sustainable and efficient.
“This work comes at a time when there are significant advancements — and potential challenges — in classical machine learning,” Jiang said. “The quantum device could help to address some of those challenges.”
Big Brains Podcast: Combating our global water crisis using AI, with Junhong Chen
In a recent episode of UChicago’s Big Brains podcast, UChicago PME Prof. Junhong Chen outlined his research using AI and machine learning to address many of the global water crises in surprising ways.
Chen also is the Lead Water Strategist at Argonne National Laboratory and co-Principal Investigator and Use-Inspired R&D Lead for Great Lakes ReNEW, which in January received a $160 million federal climate award to clean the Great Lakes’ water.
“By treating the water to the right level of purity, for the right applications, you can save energy,” Chen said. “AI and machine learning can really help accelerate that process.”
Researchers boost vaccines and immunotherapies with machine learning to drive more effective treatments
In a potential first for the field of vaccine design, machine learning was used to guide experimental testing of new immune pathway-enhancing molecules and discovered one particular small molecule that could outperform the best immunomodulators on the market. The results are published in the journal Chemical Science.
“We used artificial intelligence methods to guide a search of a huge chemical space,” said Prof. Aaron Esser-Kahn, co-author of the paper who led the experiments. “In doing so, we found molecules with record-level performance that no human would have suggested we try. We’re excited to share the blueprint for this process.”
As AI becomes more present across academia and industry, UChicago PME researchers are utilizing it in new and innovative ways. Learn more about AI at UChicago Pritzker Molecular Engineering.