The GWAS results from the MDD phenotype in 3 discovery samples (PGC MDD29, UKB and GS:SFHS) were used to build polygenic profile scores (PGS) in GS:SFHS and UKB, incorporating SNPs with a discovery sample association P-value cut-off of PT ≤ 0.01, PT ≤ 0.05, PT ≤ 0.1, PT ≤ 0.5, and all SNPs (PT ≤ 1). Results from PT ≤ 0.05 are shown in Fig. 4, as this PT explained the most variance in the target datasets. Results from all PT are shown in Supplementary Tables 18-21. PGS derived using information from the PGC MDD29 GWAS yielded significant associations between depression phenotype and PGS across almost all thresholds in both GS:SFHS and UKB (with the exception of mMDD at PT ≤ 0.01 in GS:SFHS). PGS derived using information from the UKB MDD GWAS yielded significant associations with MDD, rMDD, and mMDD phenotypes in GS:SFHS at all thresholds, however fMDD did not survive multiple testing correction at any PT. PGS derived using information from the GS:SFHS MDD GWAS yielded significant associations with MDD, rMDD, and fMDD phenotypes in UKB at all
