Pathogenic de novo SNVs, indels and CNVs in 1,000 children with undiagnosed developmental disorders. M. Hurles1, M. van Kogelenberg1, T. Fitzgerald1, W. D. Jones1, D. King1, P. Vijayarangakannan1, S. Gerety1, K. Morley1, S. Gribble1, D. Barrett1, K. Ambridge1, N. Krishnappa1, E. Prigmore1, D. Rajan1, T. Bayzetinova1, S. Al-Turki1, A. Tivey1, S. Clayton1, R. Miller1, P. Jones1, N. Carter1, C. Wright1, J. Barrett1, D. FitzPatrick2, H. Firth1,3, DDD Study 1) Wellcome Trust Sanger Inst, Cambridge, United Kingdom; 2) MRC Human Genetics Unit, MRC IGMM, University of Edinburgh, Western General Hospital, Edinburgh, UK; 3) East Anglian Medical Genetics Service, Cambridge University Hospitals NHS Foundation Trust, Cambridge Biomedical Campus, Cambridge, UK.

   To delineate the genetic architecture of severe undiagnosed developmental disorders in UK children we have deeply phenotyped over 6,000 affected children and their parents through a nationwide network of clinical geneticists, and recruited the families into a genetic research study entitled the Deciphering Developmental Disorders study. Seventy-five percent of the families are sporadic. We are interrogating the causal roles of coding and regulatory SNVs, indels and CNVs by applying exome-array comparative genomic hybridization (exome-aCGH) to detect deletions and duplications, and exome-sequencing to detect sequence variants, in all coding exons, known enhancers, and the most highly conserved non-coding elements. We have profiled over 5,000 probands using exome-aCGH, and over 1,000 parent-proband trios with exome sequencing. We are currently able to provide likely diagnoses for 15-20% of children. We have identified recurrent functional de novo mutations in 45 genes, of which only 19 are already known developmental disorder genes. The largest single contributing gene is ARID1B. We have identified four genes where exactly the same mutation occurs in 2 or more families, highly suggestive of gain-of-function mutations. In aggregation, these analyses have identified more than 10 likely novel developmental disorder genes. We have modeled some of these plausible candidate genes in zebrafish and identified concordant developmental phenotypes in morphant zebrafish for a subset of these. We will describe the breakdown of these pathogenic and putatively pathogenic variants by phenotype and family history.