Backbone flexibility in computational protein design

Abstract

The field of computational protein design has produced striking successes, including the engineering of novel enzymes. Many of these achievements employed methodologies that sample amino acid side-chains on a fixed backbone, while methods that explicitly model backbone flexibility have so far largely focused on the design of new structures rather than functions. Recent methodological improvements in conformational sampling techniques, some borrowed from the field of robotics to model mechanically accessible conformations, now provide exciting opportunities to explore amino acid sequences and backbone structures simultaneously. Incorporating functional constraints into flexible backbone design should help to achieve challenging engineering goals that exploit the role of conformational variability in controlling biological processes, while more generally advancing biophysical understanding of the relationship between variations in protein sequence, structure, dynamics, and function.