Comment on ‘protein–protein binding affinity prediction from amino acid sequence’

Predicting the strength of interactions between globular proteins is a central and important topic in structural bioinformatics (Moal et al., 2013). The amino acid sequence represents the chemical bonding in a protein which, along with the solvent, dictates how it folds into an ensemble of thermally accessible states. In turn, structure specifies the strength and identity of its binding partners, by establishing the specific arrangements of intermolecular interactions and the intramolecular strain required to achieve them.