It is important to develop consensus guidelines for defining off-target mutations in DNA, which could occur as an unintended side-effect of genome editing (see Nature 522, 20–24; 2015). I encourage the community to contribute to these discussions (see go.nature.com/zncbil).

Such uniform standards would help researchers, peer-reviewers, journal editors and regulators to best identify such mutations.

For therapeutic applications, unwanted mutations need to be defined by the most highly sensitive, unbiased genome-wide methods — given that even low-frequency events in large populations of cells could have clinical consequences. Such a comprehensive definition might not be necessary for research projects because appropriate control experiments would exclude the potentially confounding effects of off-target actions.

For now, direct comparison of state-of-the-art technologies can start to define best practices. Refinement will follow as detection and editing methodologies advance.