ORIGINAL ARTICLE A genome-wide association study in 574 schizophrenia trios using DNA pooling G Kirov 1 , I Zaharieva 2 , L Georgieva 1 , V Moskvina 1 , I Nikolov 1 , S Cichon 3 , A Hillmer 3 , D Toncheva 2 , MJ Owen 1 and MC O’Donovan 1 1 Department of Psychological Medicine, Cardiff University, Henry Wellcome Building, Heath Park, Cardiff, UK; 2 Department of Medical Genetics, University Hospital ‘Maichin Dom’, Sofia, Bulgaria and 3 Department of Genomics, Life and Brain Center, University of Bonn, Bonn, Germany The cost of genome-wide association (GWA) studies can be prohibitively high when large samples are genotyped. We conducted a GWA study on schizophrenia (SZ) and to reduce the cost, we used DNA pooling. We used a parent–offspring trios design to avoid the potential problems of population stratification. We constructed pools from 605 unaffected controls, 574 SZ patients and a third pool from all the parents of the patients. We hybridized each pool eight times on Illumina HumanHap550 arrays. We estimated the allele frequencies of each pool from the averaged intensities of the arrays. The significance level of results in the trios sample was estimated on the basis of the allele frequencies in cases and non-transmitted pseudocontrols, taking into account the technical variability of the data. We selected the highest ranked SNPs for individual genotyping, after excluding poorly performing SNPs and those that showed a trend in the opposite direction in the control pool. We genotyped 63 SNPs in 574 trios and analysed the results with the transmission disequilibrium test. Forty of those were significant at P < 0.05, with the best result at P = 1.2  10 6 for rs11064768. This SNP is within the gene CCDC60, a coiled-coil domain gene. The third best SNP (P = 0.00016) is rs893703, within RBP1, a candidate gene for schizophrenia. Molecular Psychiatry (2009) 14, 796–803; doi:10.1038/mp.2008.33; published online 11 March 2008 Keywords: schizophrenia; pooled DNA; pooling; Illumina; genome-wide association Introduction Schizophrenia (SZ) has a strong genetic component, as shown by heritability estimates of B80%. 1 How- ever, there is no clear mode of transmission and most cases appear to be sporadic, suggesting a complex pattern of inheritance. Like other disorders of com- plex inheritance, it is now believed that many genes of small effect operate in the aetiology of this disorder. Epidemiological and molecular genetic studies sug- gest that genetic risk is mainly attributable to multiple alleles each with a small to moderate effect on liability. Several promising candidate susceptibility genes have been reported, but most of the genetic risk has not yet been attributed to specific genes. 2 Recent technological advances have enabled re- searchers to perform genome-wide association (GWA) studies using hundreds of thousands of single- nucleotide polymorphisms (SNPs) that can capture the majority of common variation in the human genome. Several such studies have already been published for disorders such as, diabetes, ischaemic heart disease, hypertension and others (for example see ref 3 ). These studies have produced many defini- tive findings, and in many cases, implicated genes that were not expected to be involved (for example, in ischaemic heart disease). The odds ratios found were modest even for the top loci (1.18–5.49 for hetero- zygote odds ratios and 1.48–18.52 for homozygote odds ratios in the largest study on seven different common disorders 3 ) and required thousands of cases and controls to be used in the initial or the replication stages, to reach results that are genome-wide signifi- cant (that is, corrected for the number of SNPs tested). It is expected that any future susceptibility factors discovered in complex diseases will have even lower odds ratios, unlikely to be > 2.0. The need to genotype very large samples translates into very high costs that are beyond the budgets of most research groups in most countries. Although the sensitivity and specificity of pooled DNA analyses are imperfect, previous work in our own and other laboratories suggests that such analysis at the level of single locus 4 and highly parallel chip-based methods 5–8 can offer an economic alternative to individual genotyping, although clearly, not all loci with evidence for association in the samples so analysed will be detected. Several array-based Received 24 October 2007; revised 1 February 2008; accepted 8 February 2008; published online 11 March 2008 Correspondence: Dr G Kirov, Department Psychological Medicine, Cardiff University, School of Medicine, Heath Park, Henry Wellcome Building, Cardiff CF14 4XN, UK. E-mail: kirov@cardiff.ac.uk Molecular Psychiatry (2009) 14, 796–803 & 2009 Nature Publishing Group All rights reserved 1359-4184/09 $32.00 www.nature.com/mp