Genome-wide association study of schizophrenia in a Japanese population

Background Genome-wide association studies have detected a small number of weak but strongly supported schizophrenia risk alleles. Moreover, a substantial polygenic component to the disorder consisting of a large number of such alleles has been reported by the International Schizophrenia Consortium. Method We report a Japanese genome-wide association study of schizophrenia comprising 575 cases and 564 controls. We attempted to replicate 97 markers, representing a nonredundant panel of markers derived mainly from the top 150 findings, in up to three data sets totaling 1990 cases and 5389 controls. We then attempted to replicate the observation of a polygenic component to the disorder in the Japanese and to determine whether this overlaps that seen in UK populations. Results Single-locus analysis did not reveal genome-wide support for any locus in the genome-wide association study sample (best p = 6.2 × 10 ^-6) or in the complete data set in which the best supported locus was SULT6B1 (rs11895771: p = 3.7 × 10^-5 in the meta-analysis). Of loci previously supported by genome-wide association studies, we obtained in the Japanese support for NOTCH4 (rs2071287: p_meta = 5.1 × 10 ^-5). Using the approach reported by the International Schizophrenia Consortium, we replicated the observation of a polygenic component to schizophrenia within the Japanese population (p = .005). Our trans JapanUK analysis of schizophrenia also revealed a significant correlation (best p = 7.0 × 10^-5) in the polygenic component across populations. Conclusions These results indicate a shared polygenic risk of schizophrenia between Japanese and Caucasian samples, although we did not detect unequivocal evidence for a novel susceptibility gene for schizophrenia.