Genome-wide homozygous haplotype mapping to identify Autism Spectrum Disorder candidate genes. J. Casey1, T. Magelhaes2, J. Conroy1, R. Regan1, N. Shah1, R. Anney3, E. Heron3, A. Green1,4, L. Gallagher3, M. Gill3, D. Shields1, A. Vicente2, S. Ennis1, Autism Genome Project (www.autismgenome.org) 1) Health Sciences Center, University College Dublin, Dublin 4, Ireland; 2) Instituto Gulbenkian de Ciencia, Oeiras, Portugal; 3) Department of Psychiatry, Trinity College Dublin, Ireland; 4) National Centre for Medical Genetics, Ireland.

   Autism spectrum disorder (ASD) is a severe neurodevelopmental disorder of complex and heterogeneous aetiology. ASD is characterised by altered cognitive ability including impaired language and communication skills, deficits in social reciprocity and repetitive and restricted patterns of behaviour and interests. Evidence to support a strong genetic component in the aetiology of ASD stems largely from twin studies which demonstrate a much higher heritability in monozygotic (92%) compared to dizygotic (2-10%) twins. Recent studies offer support for the hypothesis that rare genetic variants contribute to the disorder. However the systematic detection of susceptibility genes containing low-frequency mutations in the genome is technically challenging. Large tracts of extended homozygosity represent a newly explored form of genetic variation and may be highly relevant in disease gene discovery. The mapping of homozygous regions of matching haplotype on a genome-wide scale could help to identify low-frequency potentially disease-liable candidate genes associated with complex disorders. A cohort of ~1500 ASD trios was genotyped for 1 million SNPs allowing adequate coverage for detection of extended runs of homozygosity. The 4000 sample cohort was clustered and separated into analysis groups based on common population ancestry. For each group we applied a whole-genome homozygous haplotype (HH) mapping approach to identify HH that are statistically more prevalent in ASD probands compared to parental controls. Such regions may facilitate the identification of novel risk genes contributing to this complex disorder. The genes within the significant HH were assessed in terms of their biological relevance to the ASD phenotype. Our research identified a number of novel candidate ASD genes of functional and positional relevance. We also replicated significant findings across multiple population analysis groups and the main findings will be presented at the ASHG 2009. Furthermore an unexpectedly high proportion of published ASD genes are located within HH that are significantly more common in ASD patients. Pathway analysis of the significant candidates showed an excess of genes involved in axon guidance, an interesting observation given the recent implication of this gene family in ASD. Our study highlights the utility of HH mapping as a candidate-gene identification strategy and may improve our understanding of the molecular network underlying this complex disorder.