Identifying Novel Chemical-Related Susceptibility Genes for Five Psychiatric Disorders Through Integrating Genome-Wide Association Study and Tissue-Specific 3′aqtl Annotation Datasets

Sirong Shi,Huijie Zhang,Xiaoge Chu,Qingqing Cai,Dan He,Xiaoyue Qin,Wenming Wei,Na Zhang,Yijing Zhao,Yumeng Jia,Feng Zhang,Yan Wen

DOI: https://doi.org/10.1007/s00406-023-01753-0

2024-01-01

European Archives of Psychiatry and Clinical Neuroscience

Abstract:The establishment of 3ʹaQTLs comprehensive database provides an opportunity to help explore the functional interpretation from the genome-wide association study (GWAS) data of psychiatric disorders. In this study, we aim to search novel susceptibility genes, pathways, and related chemicals of five psychiatric disorders via GWAS and 3ʹaQTLs datasets. The GWAS datasets of five psychiatric disorders were collected from the open platform of Psychiatric Genomics Consortium (PGC, https://www.med.unc.edu/pgc/ ) and iPSYCH ( https://ipsych.dk/ ) (Demontis et al. in Nat Genet 51(1):63–75, 2019; Grove et al. in Nat Genet 51:431–444, 2019; Genomic Dissection of Bipolar Disorder and Schizophrenia in Cell 173: 1705-1715.e1716, 2018; Mullins et al. in Nat Genet 53: 817–829; Howard et al. in Nat Neurosci 22: 343–352, 2019). The 3′untranslated region (3′UTR) alternative polyadenylation (APA) quantitative trait loci (3′aQTLs) summary datasets of 12 brain regions were obtained from another public platform ( https://wlcb.oit.uci.edu/3aQTLatlas/ ) (Cui et al. in Nucleic Acids Res 50: D39–D45, 2022). First, we aligned the GWAS-associated SNPs of psychiatric disorders and datasets of 3′aQTLs, and then, the GWAS-associated 3ʹaQTLs were identified from the overlap. Second, gene ontology (GO) and pathway analysis was applied to investigate the potential biological functions of matching genes based on the methods provided by MAGMA. Finally, chemical-related gene-set analysis (GSA) was also conducted by MAGMA to explore the potential interaction of GWAS-associated 3ʹaQTLs and multiple chemicals in the mechanism of psychiatric disorders. A number of susceptibility genes with 3ʹaQTLs were found to be associated with psychiatric disorders and some of them had brain-region specificity. For schizophrenia (SCZ), HLA-A showed associated with psychiatric disorders in all 12 brain regions, such as cerebellar hemisphere (P = 1.58 × 10–36) and cortex (P = 1.58 × 10–36). GO and pathway analysis identified several associated pathways, such as Phenylpropanoid Metabolic Process (GO:0009698, P = 6.24 × 10–7 for SCZ). Chemical-related GSA detected several chemical-related gene sets associated with psychiatric disorders. For example, gene sets of Ferulic Acid (P = 6.24 × 10–7), Morin (P = 4.47 × 10–2) and Vanillic Acid (P = 6.24 × 10–7) were found to be associated with SCZ. By integrating the functional information from 3ʹaQTLs, we identified several susceptibility genes and associated pathways especially chemical-related gene sets for five psychiatric disorders. Our results provided new insights to understand the etiology and mechanism of psychiatric disorders.