UO Chemistry and Biochemistry – Essential Elements

UO biochemistry major named finalist for Gates Cambridge Scholarship

Posted on January 28, 2026 by zoelewis

Student Amelia Kotamarti in lab

Physical Chemistry Seminar – Dhiman Ray, February 2nd

Posted on January 26, 2026January 26, 2026 by zoelewis

PChem seminar poster-Dhiman Ray, Feb. 2nd, 2026

Department of Chemistry and Biochemistry
Physical Chemistry Seminar Series

Professor Dhiman Ray, University of Oregon
February 2, 2026—2:00pm
Tykeson 140

“Learning Mechanisms from Explainable Artificial Intelligence”

Molecular dynamics (MD) simulations are widely used to study atomistic mechanisms of chemical, material, and biological processes. Although most practical systems exhibit beyond millisecond timescales (e.g., drug-target binding and protein folding), we can only simulate up nanoseconds to microseconds using all atom MD simulation. Enhanced sampling algorithms address this problem by accelerating conformational sampling using an external biasing potential. The accuracy and efficiency of these algorithms are sensitive to the choice of collective variable (CV), a low-dimensional space along which the bias is applied. Deep neural networks can be used to construct CVs using a generic and system-agnostic feature space to compute an accurate free energy surface for complex molecular processes. However, their lack of interpretability and high cost of evaluation during trajectory propagation make NN-CVs difficult to apply to biomolecular processes.
In the past 1.5 years at the University of Oregon, my research group has been addressing this limitation from two different directions. First, we developed a surrogate model approach to express the output of a neural network as a linear combination of a subset of the input descriptors. In addition to providing mechanistic insights due to their explainable nature, the surrogate model CVs exhibits improved computational efficiency and negligible reduction in accuracy compared to the NN-CVs in reconstructing the underlying free energy surface. Second, we introduced a deep-multitask learning algorithm that can identify mechanistic pathways without the need for spending computational effort in sampling and converging the whole free energy landscape. This approach also generates mechanistic fingerprints from which subsequent ML models can be trained to predict dynamics properties of molecules. These explainable AI techniques provide mechanistic understanding of RNA and protein dynamics at an atomistic resolution.
1. Elangovan, R., Chatterjee, S., Ray, D. (2025). Data-driven enhanced sampling of mechanistic pathways. Proceedings of the National Academy of Sciences 122(49), e2517169122.
2. Chatterjee, S., & Ray, D. (2025). Acceleration with interpretability: A surrogate model-based collective variable for enhanced sampling. Journal of Chemical Theory and Computation, 21(4), 1561-1571.
3. Chatterjee, S., & Ray, D. (2026). Characterizing RNA Tetramer Conformational Landscape Using Explainable Machine Learning. The Journal of Physical Chemistry Letters.
4. Chatterjee, S., & Ray, D. (2026). Explainable Machine Learning Guided Enhanced Sampling of Protein Conformational Transition in HSP90. ChemRxiv.

Andrew Marcus article “Functional Integration of the Bacteriophage T4 DNA” selected as Front Cover issue of Biochemistry.

Posted on January 23, 2026January 26, 2026 by zoelewis

Cover art from manuscript “Functional Integration of the Bacteriophage T4 DNA”

Functional Integration of the Bacteriophage T4 DNA Replication Complex: The Multiple Roles of the ssDNA Binding Protein (gp32)

Authors: Claire S. Albrecht, Brett Israels, Jack Maurer, Peter H von Hippel, Andrew H. Marcus

Physical Chemistry Seminar – Matthias Heyden, January 26th

Posted on January 22, 2026January 26, 2026 by zoelewis

Seminar Event poster Matthais Heyden headshot, event info and abstract

Department of Chemistry and Biochemistry
Physical Chemistry Seminar Series

Professor Matthias Heyden, Arizona State University
January 26, 2026—2:00pm
Tykeson 140

“Pushing Proteins in the Right Direction”

Protein function often depends on dynamic transitions between conformations, whose relative stability can be modified by intermolecular interactions, e.g., upon ligand binding, formation of multimeric protein complexes, in crowded environments, etc. This creates several pathways to modulating protein activity and function.

However, conformational transitions in proteins are rare events that can be challenging to characterize experimentally or to sample in unbiased all-atom simulations. Therefore, we developed an enhanced sampling protocol for molecular dynamics simulations that can be applied to folded proteins without requiring system-specific information.

The key ingredients for the most effective enhanced sampling strategies are so-called collective variables (CVs), which ideally describe a low-dimensional space that is closely aligned with the conformational transitions of the protein. The challenge is that these CVs need to be known prior to the start of the simulation.

We demonstrate for multiple protein systems that suitable CVs can be extracted from fluctuations in short unbiased simulation trajectories that can be generated at little cost. Our main innovation is an analysis of protein vibrations that successfully isolates even the most anharmonic low-frequency vibrations. Of particularly importance are collective degrees of freedom that fluctuate at zero frequencies and thus do not describe simple oscillations. In addition to rigid-body protein diffusion, the zero-frequency dynamics are dominated by damped low-frequency vibrations. The damping term is responsible for the zero-frequency response and the result of strong coupling to the surrounding solvent. These damped vibrations are excellent CVs for enhanced sampling for two reasons: 1) they modulate the shape of the protein and its interactions with the solvent (à solvent damping); and 2) they experience minimal restraining forces (à low vibrational frequency).

Our simulation protocol allows us to characterize conformational transitions and free energy surfaces for proteins with high statistical accuracy in a short time (<24h with 20 GPUs). This allows us to analyze even small changes in the relative stability of distinct protein conformations (order of kT) as a function of molecular environments (crowding or ligand binding), sequence (mutants and evolved variants), or other physical parameters such temperature and pressure.

This provides opportunities for the development of improved machine-learning models for protein conformational ensembles, the design of highly active artificial enzymes, as well as the development of selective drugs that exploit allosteric mechanisms.

Physical Chemistry Seminar – Marina Guenza, January 12th

Posted on January 5, 2026January 26, 2026 by zoelewis

P-Chem Seminar Poster - January 12, 2026

Department of Chemistry and Biochemistry
Physical Chemistry Seminar Series

Professor Marina Guenza, University of Oregon
January 12, 2026—2:00pm
Tykeson 140

“Anomalous Dynamics in Macromolecular Fluids”

While molecular motion is diffusive at long times, many complex liquids—including polymer melts, biomolecular coacervates, and systems approaching the glass transition—exhibit anomalous subdiffusive dynamics at intermediate times. This regime reflects cooperative motion arising from transient correlations between intra- and intermolecular degrees of freedom.

In this talk, I will present a theoretical framework developed in the Guenza group to describe the cooperative motion of many interacting chains. The approach quantitatively predicts experimentally accessible dynamical observables and has been validated by Neutron Spin Echo measurements performed by Dieter Richter and coworkers at the Jülich Centre for Neutron Science. Beyond polymer liquids, this framework provides a foundation for generalized Langevin equation–based coarse-grained models of coacervates and glass-forming systems, enabling predictive descriptions of collective dynamics across molecular and mesoscale regimes.