Twenty-five schizophrenia candidate genes were checked for association with schizophrenia in a cohort of 879 cases and 864 controls.

Displayed are (1) gene symbol; (2) previously associated candidate SNPs in that gene, included in the Illumina 300K chip; (3) previously associated candidate SNPs in that gene, not included in the Illumina 300K chip but can be represented by a LD proxy in the chip; (4) the closest proxy SNP present on the Illumina 300K chip for column 3; (5) the lowest p value for SNP-specific (column 2 and 4) replication; (6) Maximum odds ratio (allelic OR under multiplicative genetic model) at the previously-implicated locus, based on observed p value at this SNP or a proxy. A real odds ratio of this size is expected to produce a p value as high as the one observed only 5% of the time, based on the power formulae of Chapman et al[59]; For detailed review on the previous reports of these candidate genes and SNPs see reference [30]; (7) lowest p value of all the HumanHap300 SNPs located in that gene (lowest p); (8) number of SNPs tested in each gene; (9) the lowest p value corrected for all the SNPs tested in that gene (gene-wide correction); (10) the lowest p value corrected for all the SNPs in the 25 candidate genes (set-wide correction); (11) the lowest p value after correction for all HumanHap300 SNPs tested (genome-wide correction).