Research

In order to probe the conformational dynamics of biomolecules across biologically relevant length and time-scales, we develop coarse-grained (CG) models for proteins, nucleic acids, as well as their complexes. Recently, we introduced the Three Interaction Site (TIS) model for DNA, and the generalized COFFEE framework for meso-scale simulations. Recently, we have shown that COFFEE accurately captures salt-induced unwrapping of mononucleosomes (see figure on the left). 

Please check out our publications in JCTC (2018 and 2024).

Using the self-organized polymer model for IDPs (SOP-IDP), we look for cryptic signatures within monomer conformational ensembles (MCEs) that could explain why some sequences aggregate faster than others.  For many amyloidogenic sequences, such as the Ab peptides (see figure on left), and the FUS low complexity domains, we have shown that much can be learnt from a careful analysis of the MCEs.

Check out our latest works in PNAS (2020), JPCL (2021) and Science Advances (2023). 

Bimolecular condensates (BMCs) are highly dynamic entities, and are primarily assemblies of intrinsically disordered proteins (IDPs) and non-coding RNA (ncRNA) molecules. BMCs are believed to form via a thermodynamically driven process known as liquid-liquid phase separation (LLPS) (other mechanisms almost certainly do exist!). We are interested in finding out if configurations relevant to LLPS are pre-selected from a large pool of structures. In other words do connections exist between LLPS and the topography of the monomer energy landscapes?

Some fascinating papers on these topic are: 

Relation between single-molecule properties and phase behavior of intrinsically disordered proteins 

Sequence-dependent material properties of biomolecular condensates and their relation to dilute phase conformations 

Prediction of phase separation propensities of disordered proteins from sequence 

We exploit the framework of elastic network models (ENMs) in combination with the structural perturbation method (SPM) to understand the link between structure and function. One system of interest to us is cytochrome P450, a metalloenzyme that catalyzes key chemical reactions within the cellular machinery. Traditionally substrate binding to P450 has been studied using atomistic molecular dynamics simulations. In contrast, our approach is rather reductionist, and we are using ENMs and SPM to map out the network of residues that are critical for function. 

Is chromatin liquid-like or solid-like? For an excellent discussion on this topic please see: Chromatin: Liquid or Solid?  

Other important references: 

 Organization of Chromatin by Intrinsic and Regulated Phase Separation 

Condensed Chromatin Behaves like a Solid on the Mesoscale In Vitro and in Living Cells  

We seek to answer these and other related questions by zooming into the chromatin structure with our recently developed COFFEE framework, which provides a highly accurate description of sequence-dependent DNA-protein interactions.