Integration of Alzheimer’s Disease GWAS with Molecular QTLs Reveals Prioritized Risk Genes and Disease‐associated Molecular Alterations

Abstract Background Despite the success of genome‐wide association studies (GWAS) to identify genetic variants associated with complex diseases, as the majority of these variants reside in non‐coding genome, one important task is to link these variants to disease‐associated molecular alterations in risk genes. To this end, we developed a systematic gene prioritization methodology that integrated our recent Alzheimer’s disease (AD) GWAS meta‐analysis with molecular quantitative trait loci (QTL) in disease‐relevant tissues and identified highly likely candidate risk genes and molecular alterations within the majority of 42 novel loci identified. In this subsequent study, we updated our pipeline with new molecular QTL catalogues and analyzed the complete landscape of AD risk. Method We performed GWAS‐QTL integration in AD‐relevant tissues and cells by querying the lead variants in each locus in the molecular QTL catalogues, (ii) investigating genetic colocalization between GWAS and molecular QTL signals using coloc, and (iii) conducting transcriptome, methylome, and proteome‐wide association studies (TWAS/MWAS/PWAS) using the EADB stage I meta‐analysis GWAS on 85,934 cases and 401,577 controls. The updated gene prioritization pipeline included new analyses based on new brain protein expression QTL (pQTL), cell‐type‐specific expression QTL (ct‐eQTL), expression QTL (eQTL), histone acetylation QTL (haQTL), and methylation QTL (mQTL) catalogues. Moreover, we used Nanopore RNA sequencing to further investigate risk‐associated splicing events revealed by splicing QTL (sQTL) integration. Result Our brain ct‐eQTL coloc results showed that while the most risk‐associated gene expression alterations are specific to microglia, the proportion of astrocyte‐specific eQTL colocalizations were doubled compared to the previous studies. These included EGFR, a previously prioritized gene due to bulk brain eQTL coloc hits, whose downregulation was regulated by the protective signal specifically in astrocytes. Importantly, in a previously unresolved locus we could prioritize HS3ST5 through astrocyte‐specific eQTLs as well. Furthermore, pQTL coloc and PWAS prioritized SNX1, a previously prioritized gene at a lower confidence together with neighboring FAM96A gene, by implicating its genetic downregulation with increased AD risk. Conclusion Our updated pipeline systematically prioritized AD risk genes and disease‐associated molecular alterations in the majority of risk loci, and contributed to better interpretation of AD GWAS and understanding of genetically‐associated molecular mechanisms underlying AD risk.