Designing ensembles in conformational and sequence space to characterize and engineer proteins.

Computational protein design – the search for amino acids that adopt defined structures and functions – has achieved significant successes in recent years. Design is often approached from the perspective of targeting the design objective one structure-sequence pair at a time. Here we summarize studies that take an alternate approach, considering design solutions from the perspective of conformational and sequence ensembles. We argue that this strategy both more closely mimics the properties of naturally occurring proteins and drives progress in engineering. We discuss benchmarks for evaluating design methods, highlight recent advances, and project how the evolving capabilities of gene sequencing, gene synthesis, and genome editing may change the design field.