At neurology academic meetings at the end of the last century, neurologists would joke that in the future patients would have MR scans of their brain and spine and an analysis of their whole genome, and so neurology would be easy. How we laughed at this absurd fantasy. However, it turns out we were half right—MR imaging and whole-genome sequencing are now widely available. But if anything neurology has become more complicated. Whole-genome sequencing has been an astonishing advance in diagnostics, though like all investigations, it must be understood to get the most out of it. Christopher Record and Mary Reilly describe the lessons and pitfalls of whole-genome sequencing (page 263), highlighting where is it extremely reliable (single nucleotide variations), where reliability is improving (copy number variants and small repeat expansions) and where things can be missed or misinterpreted (for example, large repeat expansions and balanced structural variants). They use Charcot-Marie-Tooth (CMT) …