passed Bonferroni’s correction (P < 1.43 × 10−7), from which 186 also co-localized between the eQTL and GWAS traits when using coloc28 (posterior probability for co-localization of significant signals PP4 > 0.7; Fig. 5a and Supplementary Fig. 21), confirming that the two traits shared the same causal SNP. Of the prioritized findings, 282 (82 of which co-localized) were associated with the risk for 31 prioritized neurological and neuropsychiatric diseases (Table 1). We focus on multiple sclerosis (MS) and highlight two examples where MR and co-localization point to probable causal GWAS genes. For other traits, see Supplementary Note, Supplementary Fig. 22 and Supplementary Tables 11–16.Table 1Prioritized genes from the MR analysis on MetaBrain eQTLs for brain-related outcomesOutcomeGeneSNPWR (SE)PCTOutcomeGeneSNPWR (SE)PCTADCR1rs6795150.15 (0.01)1.40 x 10-23OLIMSZNF746rs1046140-0.57 (0.11)8.29 x 10-08ADSLC39A13rs3740688-0.22 (0.03)1.13 x 10-10MSMAST3rs1121888740.56 (0.10)9.74 x 10-08ADTSPAN14rs19026600.12 (0.02)4.73 x 10-09MSMYNNrs9866116-0.43 (0.08)1.22 x 10-07ADAPH1Brs1176180170.17 (0.03)1.05 x 10-08PDKANSL1rs199451-0.25 (0.03)3.35 x 10-19EXADPRSS36rs78924645-0.09 (0.02)1.63 x 10-08PDCD38rs4698412-0.24 (0.03)6.99 x 10-14ASTADINPP5Drs75695980.32 (0.06)1.74 x 10-08PDHSD3B7rs11150600-0.46 (0.07)1.90 x 10-10MICADZNF668rs23596120.26 (0.05)1.00 x 10-07PDSETD1Ars35733741-0.67 (0.11)2.43 x 10-09EXADACErs4291-0.10 (0.02)1.39 x 10-07PDRAB29rs7087230.21 (0.04)1.03 x 10-08ALSSCFD1rs2292430.17 (0.02)5.56 x 10-15PDSCARB2rs76970730.29 (0.05)6.54 x 10-08ALSG2E3rs229244-0.24 (0.03)4.23 x 10-13SCZPPP1R18rs92659540.48 (0.05)1.11 x 10-19ALSMOBPrs67720370.29 (0.05)1.10 x 10-08SCZHIST1H4Krs132172850.63 (0.07)1.40 x 10-18BDDCLK3rs9834970-0.32 (0.04)4.79 x 10-14SCZMICBrs204999-0.37 (0.05)1.82 x 10-15BDHAPLN4rs172160410.21 (0.04)4.44 x 10-09SCZFTCDNL1rs2949006-0.16 (0.02)7.97 x 10-14OLIBDGNL3rs76467410.19 (0.03)1.07 x 10-08SCZGPANK1rs77736680.24 (0.03)2.56
