SA6: COPY NUMBER VARIANTS IN BRAIN-RELATED GENES ARE ASSOCIATED WITH NEUROPSYCHIATRIC TRAITS IN CHILDHOOD

Abstract

Background Rare Copy Number Variants (CNVs) likely play an important role in common childhood neuropsychiatric disorders such as ADHD and OCD. These disorders represent the extremes of broadly distributed behavioural traits that are influenced by underlying variation in fundamental cognitive processes such as response inhibition. Using behavioural and cognitive trait-based approaches in population samples can help reduce heterogeneity and improve power to detect rare CNVs associated with these disorders. Currently we do not know the prevalence of CNVs related to neuropsychiatric disorders in youth in the general population. Using the Spit for Science sample (Toronto, Canada), we examined the association of rare CNVs with ADHD, OCD and response inhibition traits. Methods We collected DNA and quantitative trait measures from 16,718 youth (ages 6–17 years) at a local science museum. We measured ADHD traits using the Strengths and Weaknesses of ADHD and Normal Behavior (SWAN) scale, obsessive-compulsive traits using the Toronto Obsessive-Compulsive scale (TOCS) and response inhibition, a putative endophenotype for ADHD, using the Stop-Task. We genotyped unrelated Caucasians (n=5,320) using Illumina HumanCoreExome beadchips. CNVs were called using three algorithms: iPattern (ipn), PennCNV (pcnv), and QuantiSNP (qsnp) and a high confidence dataset was generated by retaining variants detected by two or more algorithms and at least 10\xa0kb in size. We defined rare variants as those with less than 0.1% frequency against our population controls. Using linear regression, we examined whether CNV burden, overall and within psychiatrically relevant gene sets (e.g., brain expression, synaptic function), is associated with response inhibition, ADHD and OCD traits. Results After stringent quality control, 4815 (90.5%) of participants remained with a total of 9,490 rare CNVs (1.97/individual on average). We observed CNVs affecting 23 genomic loci previously implicated in neurodevelopmental conditions (e.g. 22q11.2) in 59 (1.2%) of participants. Other large CNVs affecting known psychiatric genes (e.g., ASTN2, NRXN1) were also identified. Significantly more deletions (>500\xa0kb; p=0.04) and enrichment of deletions in genes involved in synapse structure, neuron projection and neurophenotypes in mice (p Discussion In a large pediatric general population sample, brain-related CNVs were associated with neuropsychiatric traits in children although the pattern of enrichment differed between ADHD, OCD and response inhibition. Our results replicate several previous CNV burden findings for ADHD and OCD suggesting a similar genetic architecture between traits and disorders. Our approach shows the feasibility of CNV analysis and quality of CNV data from the HumanCoreExome array in a large population based sample.