The genome-wide typed (“SGENE-plus”; 2,663 cases and 13,498 controls) and meta-analysis (“SGENE-plus+ISC+MGS”) samples (in total, 7,946 cases and 19,036 controls) used here were identical to those used in our previous schizophrenia GWA study and meta-analysis15 . The primary psychosis follow-up samples employed consisted of follow-up samples from our previous GWA follow-up study (9,246 schizophrenia cases and 22,356 controls)14, plus an additional 9,337 psychosis cases (1,535 schizophrenia, 7,469 bipolar disorder, 333 related psychoses) and 46,968 controls/unaffected family members. The primary follow-up samples were genotyped or imputed for all follow-up markers. The secondary follow-up samples consisted of 1,014 cases and 1,144 controls from the Göttingen Research Association for Schizophrenia (GRAS)34, 35 study. These samples, which also had been used for secondary follow-up in our previous GWA follow-up study14, were genotyped for SNPs that were genome-wide significant in the combined meta-analysis and primary follow-up samples. Table 1 summarizes the schizophrenia and psychosis datasets used in previous and current work, and Supplementary Table 1 includes details on the individual study groups. The autism samples (3,672 cases, 16,103 controls, 4,206 family members) derived from AGP,
