ARTIFICIAL CHEMICAL INTELLIGENCE @
TIWARY LAB, UNIVERSITY OF MARYLAND
I am an Associate Professor at the University of Maryland, College Park. I have a joint position in the Department of Chemistry and Biochemistry and the Institute for Physical Science and Technology. I am also an affiliated faculty member of the Chemical Physics program, Biophysics program, Applied Math & Statistics & Scientific Computation program and the Department of Materials Science and Engineering. I am an Associate Editor at the Journal of Chemical Theory and Computation, and a member of the Editorial board for Proteins. I am also a member of the Scientific Advisory Board of Schrödinger, Inc.
My lab is interested in problems at the intersection of statistical mechanics (and more generally physical chemistry), molecular simulations and Artificial Intelligence — aka Artificial Chemical Intelligence. We develop new simulation methods that can answer the following and more questions no individual way of thinking could answer on its own. These include:
How, when and why does a drug molecule stop working?
How, when and why do a bunch of atoms dancing around randomly suddenly decide to arrange themselves in beautiful crystals?
How does a protein decide when to fold and when to misfold?
Why are RNA molecules so flexible and can we predict/use their flexibility to design new non-toxic, specialized medicine?
These questions have immense health, engineering and societal ramifications. Answering them could lead to the next super-drug or super-material with targeted, cost-effective applications and minimal unwanted side-effects. However, they have a common underlying theme of rare events - that is, processes so slow that to study them on the best supercomputers would take almost the age of the universe. To answer these and other seemingly diverse looking questions we develop the next generation of molecular simulation tools that are deeply integrated with statistical mechanics and artificial intelligence, and made available to the broad scientific community in an open-source manner (GitHub).
MORE ABOUT PRATYUSH TIWARY AND THE LAB:
Google Scholar has what we have published
Pratyush's up-to-date CV (short form 2-page) (long form multi-page) has funding, teaching, outreach, awards etc.
Play with our codes at group GitHub
Check out our group members here
I am a co-organizer of the Informal Statistical Physics seminar series with Chris Jarzynski and John Weeks. Please email me if you have suggestions for speakers! In past years I have been the organizer of the Physical Chemistry seminar series.
Our science is sustained through the generous support of following organizations. If you’re interested in supporting our research, please consider making a gift through UMD Foundation.
(Pinned): We are accepting Postdoctoral Applicants through the IPST Postdoctoral Fellowship in Artificial Chemical Intelligence (deadline Feb 28, 2023).
(Pinned): Maria Cameron (UMD Math) and Pratyush are co-organizing Brin Mathematics Research Centre Workshop on "Rare Events: Analysis, Numerics, and Applications". We have limited space for UMD students/faculty to attend. Please contact either organizer for details.
2023
Jan 25: Pratyush was recognized as the Joseph O. Hirschfelder Distinguished Visitor by the University of Wisconsin-Madison.
2022
Dec 1: Our manuscript "Path sampling of recurrent neural networks by incorporating known physics" has been featured as an Editors' Highlight at Nature Communications.
Oct 10: We welcome new PhD students Suemin Lee and Vanessa Meraz for joining our group for PhDs in biophysics and chemical physics respectively.
Sep 15: We are grateful to NSF-ACCESS for awarding us around 6 million hours of high-performance computing time.
May 27: We are grateful to the NCI-UMD partnership for funding two of our seed grant proposals in collaboration with NCI labs to develop / use methods to model Tcell-antigen interactions and disover druggable human RNAs.
May 20: Molecular Dynamics simulations with help of machine learning explain how hard-to-detect drug resistance for leukemia medication is caused by a new dissociation pathway where the drug leaves three times faster. Our paper on this has been just published in Angewandte Chemie and highlighted as a Hot Paper. See associated press release from the university.
TEACHING:
In spring of 2023 I am teaching the 3-credit course CHEM 687 "Statistical Mechanics and Chemistry". I also taught it in springs of 2018, 2019, 2020 and 2021.
In fall of 2022, I taught the 3-credit course CHEM 684 "Chemical Thermodynamics".
In falls of 2018, 2020 and 2021 I taught the 3-credit course CHEM 481 "Physical Chemistry I".
In winters of 2018, 2019, 2020 and 2021, I contributed to teaching BCHM 677 "Computational Tools in Biochemistry".
Please see the teaching page for more details.