In an attempt to identify novel genes responsible for autism, a first de novo p.Lys408Valfs*31 mutation in the activity-dependent neuroprotective protein (ADNP) gene was identified in a large cohort of autistic patients [O’Roak et al., 2012b]. As sequencing of the 209 families in this cohort did not reveal a second mutation in ADNP nor hardly in any other candidate gene, a large scale resequencing study of the most promising candidates, including ADNP, was initiated using the molecular inversion probe (MIP) sequencing technology [O’Roak et al., 2012a]. In this study of 2,446 probands, an additional p. Tyr719*de novo ADNP mutation was identified. However, a single p.Q361* mutation in a patient of the NHLBI GO Exome Sequencing Project (ESP) cohort, consisting of patients with non-neurological disorders affecting heart, lung, and blood was present, and it was concluded that statistical evidence was not sufficient to prove causality of the mutations [ESP]. Confusingly, a p. Gly1094Profs*5 mutation, inherited from an unaffected parent, was reported in this same large scale targeted resequencing study. While this finding seemed to argue against causality at first sight, this
