Because our approach explores local sampling, we can identify subregions within disordered proteins that are poorly sampled, enabling a per-residue assessment of sampling. This is especially useful for assessing if we should trust conclusions from a particular subregion.

By comparing sampling among independent simulation replicas and with a well-sampled polymer model, we can distinguish between situations in which local regions fold (repeatedly adopt the same local structure) or become energetically trapped in different locally frozen states.

To address this problem, and inspired by prior work from Lyle et al. (doi.org/10.1063/1.48...), we devise an approach to place an approximate empirical upper bound on the local conformational sampling given the size of the ensemble.

One challenge with biophysical simulations of disordered proteins is that they have heterogeneous isoenergetic conformational landscapes.

Intrinsically disordered proteins are ubiquitous across the Tree of Life and play many different important roles in cellular information processing and signaling. For a more detailed review please see: www.nature.com/articles/s41...