Multi-Omics Analysis for Identifying Glial Dysfunction-Associated Genes as Therapeutic Targets and Drug Candidates in...

Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by cognitive decline, amyloid-β plaques, and neurofibrillary tangles. It's underlying mechanisms remain unclear and no disease-modifying therapies are currently available. Recent evidence has highlighted glial function as crucial factors in AD pathogenesis. Therefore, in this study, we aimed to find key AD-relevant cell populations and genes involved in glial function as well as potential therapeutic compounds. Single-cell transcriptomic data (GSE157827) and GWAS summary statistics (ieu-b-5067) were integrated using scPagwas to identify AD-associated cell subtypes and genes. Machine learning and ROC analyses were applied to refine and validate key genes, which were used to construct a diagnostic nomogram. Gene set enrichment, immune infiltration, and drug prediction analyses were performed, and the single-cell expression and pseudotime trajectories of key genes were further examined. Through scPagwas analysis, we identified oligodendrocytes and astrocytes as the major cell types associated with AD genetic risk, and revealed 27 candidate genes enriched in glial development and function, among which QKI and ZBTB20 ranked highest. Machine learning algorithms and validation across multiple independent datasets further identified QKI, ZBTB20, and C10orf90 as the key candidate genes. Single-cell analyses confirmed high expression of these genes in oligodendrocytes and astrocytes, and their dynamic, stage-associated expression patterns along glial differentiation trajectories. Drug-gene interaction and molecular docking analyses identified retinoic acid as a potential therapeutic compound targeting QKI and ZBTB20. Our findings highlighted that oligodendrocytes and astrocytes play a key role in the development of AD and provide new perspectives for future therapeutic strategies.