Predicting gene regulatory networks from multi‐omics to link genetic risk variants and neuroimmunology to Alzheimer’s disease phenotypes

Abstract

AbstractBackgroundGenome‐wide association studies have found many genetic risk variants associated with Alzheimer’s disease (AD). However, how these risk variants affect deeper phenotypes such as disease progression and immune response remains elusive. Also, our understanding of cellular and molecular mechanisms from disease SNPs to various phenotypes is still limited. To address these problems, we performed an integrative multi‐omics analysis of genotype, transcriptomics, and epigenomics for revealing gene regulatory mechanisms from disease variants to AD phenotypes.MethodFirst, given population gene expression data of a cohort, we construct and cluster its gene co‐expression network to identify modules for various AD phenotypes. Next, we predict transcription factors (TFs) regulating co‐expressed genes and SNPs interrupting TF binding sites on regulatory elements. Finally, we construct a gene regulatory network (GRN) linking SNPs, interrupted TFs, and regulatory elements to target genes and modules for AD phenotypes. This network provides systematic insights into gene regulatory mechanisms from SNPs to phenotypes. We looked at our GRNs relating to genes from shared AD‐Covid pathways (e.g. NFKB Pathway) and used machine learning to prioritize those genes for predicting Covid‐19 severityResultOur analysis predicted cross‐region‐conserved and region‐specific GRNs in 3 regions Hippocampus, Dorsolateral Prefrontal Cortex (DLPFC), Lateral Temporal Lobe (LTL). For instance, SNPs rs13404184 and rs61068452 disrupt SPI1 binding and regulation of INPP5D in Hippocampus and LTL. While rs4802200 disrupts E2F7 regulation of KCNN4 (belongs to AD LTL module), rs117863556 interrupts REST regulation of GAB2 in DLPFC. Further, we used Covid‐19 as a proxy for immune dysregulation to identify possible regulatory mechanisms for AD neuroimmunology. Decision Curve Analysis suggest our AD‐Covid genes along with linked SNPs (that outperform known genes) can be potential novel biomarkers for neuroimmunology. Finally, our results are open‐source available as a comprehensive AD functional genomic map, providing deeper mechanistic understanding of the interplay among multi‐omics, regions, gene functions, phenotypes.ConclusionOur pipeline predicts how non‐coding risk SNPs may be associated with changes in regulation and subsequent expression of genes associated with different phenotypes and pathways in AD. Moreover, we flagged 51 potential AD‐neuroinflammatory risk genes, which may be early biomarkers as neuroinflammation may begin decades before clinical onset.