Speaker
Description
Allostery—the functional coupling between distant sites in a protein—underlies biological regulation from signal transduction to metabolism and is a major target in drug discovery. The idea that allostery arises from shifts in a protein's conformational ensemble has gained broad acceptance over the past decades, but these conceptual advances have not been turned into quantitative, predictive tools.
We address this challenge for the case of two ligands binding to the same protein, where allosteric cooperativity can be captured by a single thermodynamic parameter. To this end, we sample conformational ensembles of proteins in different ligation states using atomistic molecular dynamics (MD) simulations, quantify probability densities in conformational space, and extract the cooperativity parameter through information-theoretic inference. The approach is model-free and general, works with equilibrium MD trajectories, and requires neither enhanced sampling nor predefined reaction coordinates.
I will present the underlying statistical-mechanical framework, its computational implementation, and its validation against experimentally well-characterized allosteric proteins. Beyond these specific systems, simulation-based prediction may open new avenues for understanding and exploiting allostery in drug design.