Frontiers of Computational Science

Abstract

{M}any cellular functions rely on interactions among proteins and between proteins and nucleic acids. {O}ur understanding of the principles that govern protein folding has been advanced in the recent years using the energy landscape theory and thanks to tight collaborations between experimentalists and theoreticians. {I}t is likely that our current understanding of protein folding can be applied to understand more complex cellular self-organization processes. {T}he limited success of binding predictions may suggest that the physical and chemical principles of protein binding have to be revisited to correctly capture the essence of protein recognition. {I}n this review, we discuss the power of reduced models to study the physics of protein assembly. {S}ince energetic frustration is sufficiently small, native topology-based models, which correspond to perfectly unfrustrated energy landscapes, have shown that binding mechanisms are robust and governed primarily by the protein's native topology. {T}hese models impressively capture many of the binding characteristics found in experiments and highlights the fundamental role of flexibility in binding. {T}he essential role of solvent molecules and electrostatic interactions in binding is also discussed. {D}espite the success of the minimally frustrated models to describe the dynamics and mechanisms of binding, the actual degree of frustration has to be explored to quantify the capacity of a protein to bind specifically to other proteins. {W}e have found that introducing mutations can significantly reduce specificity by introducing an additional binding mode. {D}eciphering and quantifying the key ingredients for biological self-assembly is invaluable to reading out genomic sequences and understanding cellular interaction networks.