Recurrently mutated genes contribute to the risk for developing sporadic autism spectrum disorder. B. J. O'Roak1, E. A. Boyle1, K. T. Witherspoon1, B. Martin1, C. Lee1, L. Vives1, E. Karakoc1, J. Hiatt1, D. A. Nickerson1, R. Bernier2, J. Shendure1, E. E. Eichler1,3 1) Department of Genome Sciences, University of Washington School of Medicine, Seattle, WA; 2) Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle, WA; 3) Howard Hughes Medical Institute, Seattle, WA.

   Recent exome sequencing of families with simplex or sporadic autism spectrum disorder (ASD) has highlighted the importance of de novo mutations and led to the discovery of novel autism candidate genes. However, most genes are only disrupted once in >1,000 of these simplex trios (single affected child-two unaffected parents). To address the problem of locus heterogeneity, we have optimized an ultra-low-cost (amortized reagent costs <$1/sample/gene) resequencing method based on modified molecular inversion probes (MIPs) and a statistical framework for evaluating the likelihood of recurrent mutations at individual genes (ORoak et al., Science, 2012). We initially applied this approach to a set of 44 genes, demonstrating that six are recurrently mutated in ~1% (24/2,573) of simplex ASD probands. Here we extend this paradigm further by screening 64 genes (202 kilobases/sample) in >4,000 ASD probands and >2,200 unaffected siblings. These candidate genes were selected based on de novo mutations discovered in the exomes of 1,100 probands with ASD or intellectual disability442 sequenced by us, including unpublished results from an additional 233 ASD families (932 exomes). In this study design, we enriched for genes with the most severe events, network association, chromatin function, recurrence in exome data, or novel candidates from our unpublished data. We excluded genes likely to be highly mutable or shown to carry an excess of rare severe mutations in the Exome Sequencing Project database. We made a number of protocol improvements allowing for enhanced design metrics, molecular tagging (Hiatt et al., Genome Research, 2013), and a 35% increase in target capture and evaluation (without increasing reads/sample). On average, >95% of the target regions are uniquely captured 10 or more times, including many regions that are poorly captured using current exome protocols. Preliminary results show severe mutations in ~3% of ASD probands and implicate several novel risk loci, including CACNA1D, NCKAP1, SETDB2, and a second gene in the Downs syndrome critical region. Our data strongly support a major role for recurrently disrupted genes in sporadic ASD, implicate a -catenin/chromatin remodeling network in its etiology, and provide a model for discovering and rigorously validating bona fide genetic risk loci for sporadic, genetically heterogeneous neuropsychiatric disorders.

You may contact the first author (during and after the meeting) at