Sir

Jeffrey Craig and colleagues, in Correspondence (“Harry Potter and the recessive allele” Nature 436, 776; 2005), recommend the use of analogies as tools for introducing young people to scientific concepts. Taking their example from J. K. Rowling's stories about the young wizard Harry Potter, they suggest that wizarding is a monogenic trait, with the wizard allele (W) recessive to the muggle allele (M). We believe the assumption that wizarding has a genetic basis to be deterministic and unsupported by available evidence.

Following Craig and colleagues' analogy, Hermione, as a muggle-born witch, must have WM parents. However, as Rowling fans could point out, Hermione's parents were muggle dentists who lack any family history of wizarding. It's true, of course, that chance may not have thrown up a witch or wizard for many generations, or that any who did have magical powers may have kept them secret to avoid a witch hunt.

What about Neville's apparently poor wizarding skills? These cannot be explained by incomplete penetrance, as Craig and colleagues suggest. In incomplete penetrance, individuals either display the trait or not: they do not display an intermediate degree of the trait. Poor wizarding skills might be indicative of variable expressivity of an allele. However, both variable expressivity and incomplete penetrance are associated with dominant alleles. If the wizarding allele were dominant, rather than recessive as suggested, wizarding children such as Hermione could not be born to non-wizarding parents.

Neville's clumsiness may, perhaps, be an individual characteristic unrelated to his potential powers. However, it is not possible, from the evidence presented so far, to conclude that wizarding is a heritable trait.