Decoding Alzheimer’s Molecular Signatures through Bioinformatics and AI/ML-Assisted Structure-Based Discovery of SIRT2 Inhibitors for Alzheimer’s Disease

Alzheimer’s disease (AD) is one of the neurodegenerative disorder with complicated pathogenesis. The present study aimed to explore key pathways and genes in AD pathogenesis, which could be potential targets for novel AD treatments. Differentially expressed genes (DEGs) of RNA-sequencing dataset (GSE276756) obtained from Gene Expression Omnibus (GEO) were identified using the DESeq2 R bioconductor tool. Gene ontology (GO) and pathway enrichment analyses were performed. Subsequently, a protein–protein interaction (PPI) network was constructed and analyzed as well as modules were isolated from PPI network to identify hub genes. Then, the microRNAs (miRNAs), transcription factors (TFs) and drug molecules in AD were screened out from the miRNet and NetworkAnalyst database. The PPI network, miRNA-hub gene regulatory network, TF-hub gene regulatory network and drug-hub gene interaction network were constructed by Cytoscape software. Hub genes were verified based on receiver operating characteristic (ROC) curve analysis. Finally, QSAR model development, machine learning-guided virtual screening and molecular docking study were performed for screening of novel drug molecule. A total of 958 DEGs were identified (479 up regulated and 479 down regulated genes), which were mainly enriched in terms of protein metabolic process, developmental process, aerobic respiration and respiratory electron transport, and signal transduction. Ten hub genes including GNAI1, RAC1, FANCL, NHLRC1, RNF181, PRKCA, SRC, EGFR, KMT2D and ZNRF3 were identified as potential hub genes in AD from the PPI network and its modules. MiRNAs and TFs including hsa-miR-449a, hsa-miR-129-1-3p, ELK1 and HOXA5 were identified as potential biomarkers in AD from the miRNA-hub gene regulatory network and TF-hub gene regulatory network. Isoflurophate and Phosphonotyrosine were identified as potential drugs for the treatment of AD from the DrugBank database. ROC analysis confirmed the diagnostic value of hub genes. QSAR modeling, machine learning-guided virtual screening, and molecular docking were used to identify potential inhibitors of Sirtuin 2 for AD treatment, The findings of this study provide insights into the molecular pathogenesis of AD and might provide a basis for the discovery of effective therapeutic modalities for AD.