Ilya Vakser
University of Kansas, KS·Biophysics
Proteins rarely act alone. Inside crowded cells, they bump into, bind, and assemble with one another to form molecular machines.
- Fields
Biophysics
Computer Science
Biochemistry
- Topics
protein-protein interactions
protein docking
molecular simulation
- Grants
- 3 known grants, through May 2030
What they work on
- Characterize docking-derived energy landscapes, maps of how favorable each relative position of two molecules is, for different types of macromolecules, and improve docking to capture transient interactions.
- Build simulators of crowded cellular environments that sample precomputed energy landscapes at all-atom resolution to reach extra-long timescales, and release them as public tools.
- Develop an AI-driven modeling protocol, using AI-improved energy landscapes and reinforcement learning, to simulate macromolecular assembly in cell-sized systems.
- Test whether eye lens crystallins resist self-promoting aggregation by forming native groups of varied sizes, combining AI-driven simulations with wet-lab aggregation experiments.
- Model protein interactomes of model organisms with deep learning structure prediction, and predict missense variant effects by modeling side-chain shapes and changes in interactions.
- Build GWYRE, a public resource of predicted protein complexes annotated with the effects of missense variants on phenotypes.
More about the research
Predicting a single protein's shape has become routine, but simulating how many proteins interact over long, biologically meaningful timescales remains a major challenge. The lab builds computational methods to model these interactions. It applies them to three questions: how molecules associate in the cell over long timescales, how the eye lens proteins assemble without clumping, and how genetic mutations change protein complexes.
Why it matters
Structural models of how proteins bind could give drug designers starting points by showing how existing and potential drug targets dock and associate. Linking mutations to changes in protein complexes may help researchers interpret human genetic variants. Understanding how lens crystallins avoid clumping adds knowledge of protein homeostasis and could improve how living systems are used to produce biological materials.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Modeling of macromolecular interactions in the cell
NIH · Jun 2025 – May 2030
Collaborative Research: BIO-AI: Modeling of large macromolecular systems at extra-long time scales using AI
NSF · Jul 2026 – Jun 2029
UKRI/BBSRC - NSF/BIO: Modeling of protein interactions to predict phenotypic effects of genetic mutations
NSF · Dec 2022 – Nov 2026
Closest labs
Jonah Vilseck
Indiana University-Purdue University-Indianapolis, IN·Chemistry
Use lambda dynamics free energy simulations to understand how missense mutations in a drug target cause drug resistance, and to identify drug modifications that overcome it.
Brian Kuhlman
University of North Carolina at Chapel Hill, NC·Biochemistry
Test whether a protein core mixing water-repelling and water-attracting amino acids yields a partly unfolded, water-containing cold state, using de novo designed cold unfolding proteins, biophysical measurements and molecular dynamics simulations.
Vukica Srajer
University of Chicago, IL·Biophysics
Install and test new hardware to generalize Electric-Field Stimulated X-ray crystallography, in which strong external electric fields induce motions in protein crystals that are then recorded by X-ray diffraction at atomic resolution.
The school
University of Kansas: 26 labs on PI Pages, $546M a year on research, 351 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.