Ten of these overlapping events occurred at 16p11.2, both as deletions (six) and as duplications (four). The incidence of 16p11.2 copy-number events we observe is consistent with previous studies (Mefford et al., 2009, Pinto et al., 2010 and Weiss et al., 2008). Out of the ten 16p11.2 CNVs, all but one occur in a male, and—assuming the incidence of the mutation is equal in males and females—this observation implies gender bias in penetrance of the ASD phenotype at this locus. We observe a single instance of transmission of the 16p11.2 deletion from a mother (family 12010, Table S5). Inheritance of duplications

at this locus had been previously Selleckchem GSK1210151A reported in a number of cases of ASDs, but there are relatively few reports of transmitted deletions linked to ASDs (Bijlsma et al., 2009, Fernandez et al., 2010 and Shinawi et al., 2010). A notable recurrence occurs at 7q11.23, as a duplication at the Williams-Beuren Syndrome (WBS)

locus. Deletion at this locus is associated with mental retardation, and—in contrast to ASDs—the deletion is characterized by precocious verbal ability, avid eye contact, and a highly sociable disposition (Merla et al., 2010). A third recurrent locus at 16p13.2 contains USP7 (encoding a deubiquitinase), an intriguing finding given existing links www.selleckchem.com/products/Bafilomycin-A1.html between USP7 and the spinocerebellar ataxia type 1 ( Hong et al., 2002). The fourth occurs at the NIPA locus on 15q11.2, a region in which CNVs have been previously associated with ASDs as well as epilepsy and schizophrenia ( de Kovel et al., 2010, Mefford et al., 2010, Stefansson et al., much 2008 and van der Zwaag et al., 2010). For reasons discussed later, we discount the significance of this recurrence. A fifth recurrence, de novo duplication on 16p13.11, occurs in both a proband and a sib from different families. This is a known locus of instability, wherein deletions but not duplications are thought to be associated with cognitive problems ( Hannes et al., 2009). We observe a single rare transmission of a deletion at this locus to a child with ASD (family 11450, Table

S5). To study recurrence further, we looked at a recently published study, which reported 56 de novo events in a mixed set of simplex and multiplex autism trios (Pinto et al., 2010). We omit ten of these because (1) they occur within regions that are commonly polymorphic in our cohort; and (2) when these regions are polymorphic, the polymorphisms are transmitted without bias to probands and sibs (Table S6). Of the remaining 46 events from that report, 12 events overlapped our set of de novo events in probands, at six distinct loci. Counting both data sets in total, recurrent de novo events were observed at 12 distinct loci (Table S4, Figure 5). Collections of families with multiple affected children will be significantly enriched for families at high risk of transmission. The SSC was designed to exclude such multiplex families, maximizing chances to discover de novo mutations.