Copy Number Variant Analysis Of Psychiatric Traits In A Community-Based Pediatric Sample

Abstract

Background Genetics play an important role in ADHD and OCD but only a handful of mostly common risk variants have been identified in studies that compare ADHD/OCD patients with healthy controls. Copy number variants of high penetrance may play an important role in childhood neuropsychiatric disorders. Our current knowledge of CNVs in ADHD and OCD is based on relatively small samples of patients, and to interpret the significance of these CNVs we need to assess risk CNVs in the general population. To this end we determine the frequency of clinically relevant CNVs in a large population based sample of children and adolescents and examined the association of rare CNVs with ADHD and OCD traits. Methods Quantitative data on ADHD traits using the Strengths and Weaknesses of ADHD and Normal Behavior (SWAN) scale and OC traits using the Toronto Obsessive-Compulsive scale (TOCS) was collected from 17,263 youth (ages 6-17 years) from the community. We genotyped unrelated Caucasians (n=5,366) and East Asians (n=989) using Illumina HumanCoreExome beadchips. CNV 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 10kb in size. We defined rare variants as those with less than 0.1% frequency against our population controls. We plan to examine whether CNV burden, overall and within psychiatrically relevant gene sets (e.g., brain expression, brain development), is associated with ADHD and OCD traits. We will also test if individual CNVs are associated with ADHD and OCD traits. Results After stringent quality control, 4857 (90.5%) Caucasian and 933 (94.3%) East Asian samples remained. Total number of CNVs was 31,798 with a mean size of 117,316 base pairs and a maximum size of 7.1 MB. The average number of CNVs per participants was 6. We uncovered 9,836 rare variants, of which 4,151 are deletions and 5,687 are duplications. This comprises of 7,613 rare CNVs from European subjects (mean size = 139,856 kb) and 2,225 from Asian individuals (mean size =123,141 kb). This represented a mean of 1.57 and 2.38 rare CNVs per individuals, respectively. We identified a total of 15 different known genomic disorders from 41 subjects (0.7%) in this cohort. Quantitative trait results forthcoming. Discussion Our novel approach will help elucidate the rate of psychiatrically relevant CNVs in the general population and the association of these CNVs with ADHD and OCD. Our preliminary data shows the feasibility of CNV analysis and quality of CNV data from the HumanCoreExome array in a large population based sample. Financial Relationships I (or my spouse/partner) do have potential conflicts of interest to disclose.