Identification and Validation of Blood Based Differentially Expressed Genes in Alzheimer Disease Using Integrated Transcriptomic Analysis

Abstract

Alzheimer disease is a progressive neurodegenerative disorder marked by memory impairment and cognitive decline. Early diagnosis always remain challenging because many established biomarker approaches rely on specialized imaging or invasive cerebrospinal fluid sampling, while robust blood based molecular biomarkers remain under validated Henriksen et al. (2014). To address the reproducibility gap in blood transcriptomic biomarker research, we applied a discovery and validation strategy using two independent whole blood gene expression datasets from the Gene Expression Omnibus: GSE63060 as the discovery cohort and GSE63061 as an external validation cohort Yu et al. (2021). Differential expression analysis was conducted us- ing the limma framework with multiple testing correction Ritchie et al. (2015). In the discovery cohort, differentially expressed 100 genes were identified, and 15 genes were validated consis- tently in the independent cohort. Visualization with a heatmap of top 10 DEGs, volcano plot, boxplot of a representative probe, and overlap analysis supported clear transcriptional differ- ences between control and Alzheimer disease sample. Receiver operating characteristic anal- ysis showed strong diagnostic performance for representative probes in the discovery cohort (ILMN 2097421 AUC 0.871 and ILMN 1784286 AUC 0.862), that is indicating good discrim- ination between controls and Alzheimer disease Robin et al. (2011). Overall, the overlap genes being validated provide a foundation for downstream functional studies and reproducible blood based molecular signals and clinical evaluation