Chemistry and Biochemistry’s Pratyush Tiwary Receives National Science Foundation CAREER Award
The theoretical chemist will use the award to apply artificial intelligence to the study of molecules and atoms
The University of Maryland’s Pratyush Tiwary, an assistant professor of chemistry and biochemistry with a joint appointment in the Institute for Physical Science and Technology (IPST), received a Faculty Early Career Development (CAREER) award from the National Science Foundation (NSF) for his proposal “Learning to Learn—Artificial Intelligence Augmented Chemistry for Molecular Simulations and Beyond.”
Tiwary is developing machine learning tools to model and predict thermodynamics and dynamics of complex systems at the molecular and atomic scale. He will use the $650,000 award to advance this research and to partner with researchers at Stony Brook University and the National Cancer Institute to better understand how drugs interact with molecules in the human body.
A significant outreach component of the grant will also enable Tiwary to conduct computer literacy workshops for physical science students at Prince George’s Community College and Bowie State University, as well as current and future science teachers through UMD’s Terrapin Teachers program. Students who develop an interest and aptitude for coding through the workshops will have opportunities to engage in research projects with UMD faculty members.
“I'm thrilled to be winning this award,” Tiwary said. “It reflects the importance of this kind of interdisciplinary research that I have been able to do as part of both the Department of Chemistry and Biochemistry and IPST. This award is supporting me in my efforts to tackle old problems in chemistry with really new approaches, and at the same time helping me share these new techniques with the broader community of students at UMD and elsewhere who don't have the same opportunities I’ve had.”
Using statistical mechanics, theoretical chemistry and artificial intelligence, Tiwary develops algorithms that simulate the behavior of molecules and atoms at the femtosecond timescale—that’s one millionth of one billionth of a second. Understanding the behavior of atoms at this very fine resolution will help answer questions about such things as when and how proteins fold and change shape or precisely when molecules in a given medicine stop interacting with cells in the body. This kind of work is important for advancing medicine and materials science, but it generates unmanageable amounts of data.
Tiwary is cutting through the noise of all that data to help scientists find which bits of data represent the important behaviors they need to study. To do this, he uses artificial intelligence tools that help chemists understand the exact point in time and space at which a chemical reaction occurs.
“The theme of this work is really to discover how machine learning and artificial intelligence can make chemistry better, and also how the tools of theoretical chemistry, specifically statistical mechanics, can help us understand and improve machine learning and artificial intelligence,” Tiwary said.
This new award will build on Tiwary's previous research, including his 2018 research paper titled "Reweighted autoencoded variational Bayes for enhanced sampling (RAVE)," published in the Journal of Chemical Physics. That paper, which has already been cited more than a hundred times, was one of the first to show how artificial intelligence could be used for enhancing simulations of molecular dynamics. The award will also build on Tiwary's follow-up 2019 work in the journal Nature Communications, in which he generalized the RAVE approach from 2018 through the information bottleneck approach from information theory. The bottleneck approach seeks to improve efficiency of simulations by identifying tradeoffs and striking a balance between accuracy and complexity when dealing with large amounts of data. The award will also help maintain Tiwary's commitment to releasing his work as open-source software for the broad community to use as can be seen from his GitHub page.
In 2019, Tiwary was one of 30 young researchers named to the 2019 “Future of Biochemistry” list in the journal Biochemistry. He joined UMD in 2017 after completing postdoctoral fellowships at Columbia University and the Swiss Federal Institute of Technology (ETH) in Zürich. Tiwary earned his M.S. and Ph.D. in materials science from the California Institute of Technology and a B.Tech. in metallurgical engineering from the Indian Institute of Technology in Varanasi.
The five-year CAREER awards are the NSF’s most prestigious honors in support of junior faculty members who have the potential to serve as academic role models in research and education and lead advances in the mission of their department or organization.