paperKB
coga / coga-kb
Help
Sign in

Chunk #7 — Results — Integration of GWAS and eQTL/ mQTL data from fetal and adult brain

Source
Multi-omics integration analysis identifies novel genes for alcoholism with potential overlap with neurodegenerative diseases.
Embedded
yes

Text

Summary based Mendelian Randomization (SMR) analysis of genome-wide AUD summary statistics with eQTLs and mQTLs in the adult and fetal brain identified 21 genes at 18 loci across the genome (PeQTL = 5 × 10−8; P-SMR FDR < 20%; Heidi >0.05) [Supplementary Data 4; Supplementary Figs. 7–10]. Among these 18 loci, SMR analysis nominated a single candidate causal gene at 16 loci, while more than 1 causal gene was nominated at 3p21.31 (GPX1, AMT) and 11p11.2 (SPI1, MTCH2, NUP160). To avoid the occurrence of false positive co-localizations that might be exclusively driven by stronger eQTL/mQTL signals, we focused on the loci where the strongest SNP was at least suggestively significant in the GWAS (P-GWAS < 5 × 10−5) and genome-wide significant in respective eQTL/mQTL datasets (P-eQTL/mQTL <5 × 10−8) [Table 1; Fig. 3]. Because of the much larger sample size of the DPW GWAS, 61 genes at nearly 31 loci passed the threshold for significance (P-SMR FDR < 20%; Heidi >0.05; GWAS P < 5 × 10−5; eQTL/mQTL P < 5 × 10−8) [Supplementary Data 5; Fig. 3; Supplementary Figs.