Deciphering Alzheimer's disease transcriptomics: exploration and validation of core genes in tau and Aβ pathological models toward novel therapeutic targets.

Introduction: To decode the pathology of Alzheimer's disease (AD), this study employs multi-omics approaches and bioinformatics analyses to explore AD-associated differentially expressed genes (DEGs), dissect the underlying mechanisms, and thereby facilitate the identification of core genes as well as the development of targeted therapeutic strategies. Methods: Six independent AD datasets were collected from the Gene Expression Omnibus (GEO) database, and data were processed and normalized using the R software. The evaluation of relationships between differentially expressed genes (DEGs) and AD encompassed differential expression analysis, expression quantitative trait loci (eQTL) analysis, and Mendelian randomization (MR) analysis. Additionally, gene set enrichment analysis (GSEA), immune cell correlation analysis, and Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were employed to investigate the functional roles and pathways of these genes. Machine learning approaches were applied to identify potential genes from differentially expressed genes (DEGs) associated with AD. The diagnostic performance of these candidate genes was assessed using a nomogram and receiver operating characteristic curves. The expression levels of the identified genes were further validated via quantitative real-time polymerase chain reaction (qRT-PCR). Results: Differential gene analysis identified 294 highly expressed genes and 330 lowly expressed genes, and MR analysis identified 10 significantly co-expressed genes associated with AD, specifically METTL7A, SERPINB6, VASP, ENTPD2, CXCL1, FIBP, FUCA1, TARBP1, SORCS3, and DMXL2. Noteworthy observations naive CD4+ T cells in AD, with this distinct from CIBERSORT analysis included the presence of unique immune cell subset further underscoring the critical role of immune processes in the pathogenesis and progression of the disease. METTL7A, SERPINB6, VASP, ENTPD2, FIBP, FUCA1, TARBP1, SORCS3, and DMXL2 were selected for nomogram construction and machine learning-based assessment of diagnostic value, demonstrating considerable diagnostic potential. Furthermore, the significance of the identified key genes was corroborated using both the GEO validation set and qRT-PCR. Conclusion: METTL7A, SERPINB6, VASP, ENTPD2, FIBP, FUCA1, TARBP1, SORCS3, and DMXL2 may regulate the progression of AD. These findings not only deepen our mechanistic understanding of AD pathology but also provide potential candidate genes for the development of targeted therapeutic strategies against AD.