Multiple genome wide and candidate gene analysis have been performed in order to identify genetic varaints that explain the inter-individual differences in intelligence(Butcher, L. M. et al., 2005; Butcher, L. M., Davis, O. S., Craig, I. W., & Plomin, R., 2008; Plomin, R., 1999; Plomin, R. et al., 2004; Plomin, R. et al., 2001; Buyske, S. et al., 2006; Dick, D. M. et al., 2006; Luciano, M. et al., 2006; Posthuma, D. et al., 2005, for a review see Posthuma and de Geus (Posthuma, D. & de Geus, E. J. C., 2006)) but so far results are far from clear. Of all reported genetic association studies in the literature, only 4% have shown replicable association according to a 2002 search (Hirschhorn, J. N., Lohmueller, K., Byrne, E., & Hirschhorn, K., 2002). Using data from the IMAGE project, we performed a GWAS on IQ using data from 947 European Caucasian nuclear families (2844 individuals) from eight countries (Belgium, England, Germany, Holland, Ireland, Israel, Spain, and Switzerland) that were included in the analysis. Families were recruited based on having one child with ADHD (Attention-Deficit/Hyperactivity Disorder) and another who would provide DNA and quantitative trait data. In addition, both parents had to be available for DNA-sampling. IQ scores were available for 634 probands (for which we also had genotyping data), of which 554 are males, with a mean age of 10.99 (SD 2.74). IQ was measured with WISC-III-R (Wechsler Intelligence scales for children) (Wechsler, D, 1991) or the WAIS-III-R (Wechsler Adult Intelligence Scale) (Wechsler, D, 1997) when appropriate (for children aged 17 and older). The vocabulary, similarities, picture completion and block design subtests from the WISC were used to obtain an estimate of child's IQ (prorated following procedures described by Sattler) (Sattler, J. M., 1992). Age-appropriate national population norms were available for each participating site and these were used to derive standardized estimates of intelligence (Sonuga-Barke, E. J. et al., 2008). Standardized Full Scale IQ scores ranged from 56 to 154, with a mean of 100.7 (SD 15.6).

The IMAGE consotium makes part o the Genetic Association Information Network (GAIN) a public-private partnership of FNIH (Foundation for the National Institutes of Health, Inc.) that currently involves NIH, Pfizer, Affymetrix, Perlegen Sciences, Abbott, and the Eli and the Edythe Broad Institute (of MIT and Harvard University (http,//www.fnih.org)). Genotyping was conducted at Perlegen Sciences using their genotyping platform, which comprises approximately 600,000 tagging SNPs designed to be in high linkage disequilibrium with untyped SNPs for the HapMap populations. Genotype data were cleaned by NCBI (The National Center for Biotechnology Information). Quality Control analyses were processed using the GAIN QA/QC Software Package (version 0.7.4) developed by Goncalo Abecasis and Shyam Gopalakrishnan at the University of Michigan. Details of the genotyping and data cleaning process for the ADHD GAIN study (Study Accession, phs000016.v1.p1) have been reported elsewhere (Neale, B. M. et al., 2008a). Briefly, we selected only SNPs with minor allele frequency (MAF) >= 0.05 and Hardy-Weinberg equilibrium (HWE) (P >= 1x10-6). Genotypes causing Mendelian inconsistencies were identified by PLINK and removed (http,//pngu.mgh.harvard.edu/purcell/plink/) (Purcell, S. et al., 2007a). We additionally removed SNPs that failed the quality control metrics for the other two GAIN Perlegen studies (for Major Depression Disorder (dbGAP Study Accession, phs000020.v1.p1) and Psoriasis (dbGAP Study Accession, phs000019.v1.p1). With this filtering, 384.401 SNPs were retained in the final dataset. To increase coverage in the targeted genomic areas, we used the imputation approach implemented in PLINK (v1.04), which imputes genotypes of SNPs that are not directly genotyped in the dataset, but that are present on a reference panel. The PLINK algorithm is an extension of multimarker tagging. The reference panel used consisted of 2,543,285 polymorphic autosomal SNPs genotyped on the 60 HapMap CEU founders which are publicly available for download from the HapMap website (http://www.hapmap.org). A threshold of 0.80 confidence level was set for a hard call to be included in further association testing. Most likely genotypes on imputed SNPs were then included in association analyses. Gene coverage was determined by the sum of the typed and imputed SNPs as well as the tagged SNPs divided by the total known common SNPs within a gene, using WGAviewer (Ge, D. et al., 2008). Association analysis of 634 ADHD probands was conducted using the linear procedure implemented in PLINK using the prorated IQ score as quantitative outcome and adjusting for the language in which the IQ tests were done. Initial findings show several SNPs with interesting (non geneome wide significant) association p values (p less than 10E-06) including a haplotype block on chromosome 3 (rs7643566, rs6800161, rs9860985, rs7642425, rs1602, rs480668, rs6583190 and rs553783 (uncorrected p values = 5.32E-6)) also SNP rs11594325 (uncorrected p value = 6.43 E-6) in the PRKG1 gene on chromosome 10 and SNP rs2134947 (uncorrected p value=9.47E-7) in the CNTN5 genme on chromosme 11. In order to validate these findings, an independent replication sample is needed. We propose to use the genetic and IQ information available from the ALSPAC study in order to replicate these association findings. ALSPAC, as a birth cohort is comparable to an ADHD sample as IMAGE in terms og IQ because the selection criteria applied for the IMAGE participants only required that the IQ was equal or higher than 70, meaning that the distribution of the trait in our ADHD sample is perfectly normal (overall mean IQ [all probands] = 100.0, SD=15.4). Also, the proportion of IQ values lower that this threshold is low, skewing our IQ distribution only slightly, skewness 0.103, SE=0.070), this will not influence our comparison with cohort such as ALSPAC.