Biological macromolecules are not static objects but their structures fluctuate dynamically.
In addition to thermal fluctuations, transitions between different states are especially important for biological function.
Specific systems that we have studied in detail include RNA polymerase II, the MutS DNA mismatch recognition system, kinases, and the NS3/4a viral protease.
We capture dynamics traditionally via molecular dynamics computer simulations and more recently via generative AI models.
Topics that we are currently interested in:
- Conformational sampling of highly dynamic regions
- Generation of structural ensembles consistent with experimental data, especially from NMR or cryoEM
- Rapid generation of functional states from a single input structure
Software and modeling resources:
- alphafold-multistate: state-specific modeling using AlphaFold
- idpgan: first generation generative modeling of IDPs trained on COCOMO or ABSINTH
- idpsam: second generation generative modeling of IDPs trained on ABSINTH
- asam: generative modeling of conformational ensembles
