Four papers recently reported exonic mutations in autism [4•,5•,6•,7]. Rates of non-sense mutations were elevated, though only modestly so. More autism patients were sequenced than in the schizophrenia studies, so an important question was whether mutations ever occurred in the same gene. This is a different question from that of asking whether coding mutations are more common in patients than controls. And it is an important question, because finding recurrent mutations could lead to the identification of a gene, or genes, causally implicated in the disease. Taken together, in all three studies, 18 genes with two functional mutations were found. However, by chance about 12 genes would be expected to have more than one mutation, and the enrichment was not quite statistically significant (P = 0.063) [5•].
