9 ± 1.7 and the mean nonverbal IQ was

88.4 ± 1.4. Self-reported ancestry was as follows: White non-Hispanic, 74.5%; mixed, 9.3%; Asian, 4.3%; White Hispanic, 4.0%; African-American, 3.8%; other, 4.2%. Additional phenotypic data may be found in recent publications (Fischbach and Lord, 2010) and at www.sfari.org/simons-simplex-collection. DNA samples derived from whole blood (n = 4381), cell lines (n = 68), or saliva (n = 8) were genotyped on the Illumina IMv1 (334 families) or Illumina IMv3 Duo Bead arrays (840 families), which share 1,040,853 probes in common. CNV prediction was performed by PennCNV (PN) (Wang et al., 2007), QuantiSNP (QT) (Colella et al., 2007), and GNOSIS (GN), (www.CNVision.org) (Figure 1). To assess selleck screening library detection accuracy, we evaluated 115 predicted rare CNVs (≤50% of the span of the event found at > 1% in the Database

of Genomic Variation [DGV; Cabozantinib http://projects.tcag.ca/variation/]) by quantitative polymerase chain reaction (qPCR). A higher positive predictive value was observed for CNVs called by PN and QT, with or without GN (PPV = 97% with GN, PPV = 83% without) than for other combinations of algorithms, irrespective of the number of probes mapping within the structural variation (Table S2 and Figure S1); these “high-confidence” criteria were subsequently used to identify all rare transmitted CNVs. Given a particular interest in de novo variation and the relative challenge of accurately detecting these CNVs (Lupski, 2007), we sought to optimize our detection strategy further for this class of structural variation by using the first 585 quartets with complete genotyping data (Figure 1). We identified de novo events from among the predicted rare high-confidence CNVs based on the combination of within-family intensity and genotypic data and used a blinded qPCR confirmation process (Figure S1). Fifty-three percent of de novo predictions based on ≥20 probes (n = 94) were

confirmed compared with 2.6% based on <20 probes (n = 430). Eighty-two percent of failures were false-positive predictions in offspring; 18% were false-negatives in parents. The data from this experiment were then used to further refine de novo prediction thresholds (Supplemental Experimental Procedures). In addition, given the large number of predictions of small CNVs, and the low yield of true positives in the pilot most data set (Figure S1), we elected to restrict all further statistical analysis to those rare de novo events that both encompassed ≥20 probes and were confirmed by qPCR in whole-blood DNA (Figure S1). Subsequently, at the conclusion of our study, we were able to evaluate our methods further via a comparison of confirmed de novo CNVs identified in our study versus those detected by Nimblegen 2.1M arrays from among a total of 1340 overlapping subjects (probands or siblings), as described by Levy and colleagues in this issue (Levy et al., 2011).