Basic Science and Pathogenesis

Alzheimer's disease (AD) exhibits substantial molecular heterogeneity across individuals; however, conventional cohort-based case vs. control omics analyses often mask individual-level variability in molecular disease mechanisms. Emerging single-subject omics approaches enable a more granular understanding of AD by capturing pathway-specific disruptions at the patient level. Leveraging these methods can improve disease stratification and inform personalized therapeutic strategies.

We analyzed RNA-Seq transcriptomics and proteomics data from post-mortem dorsolateral prefrontal cortex (DLPFC) in the ROSMAP cohort (913 individuals: 377 AD, 345 asymptomatic AD (asymAD), 191 controls). Individual-based gene deregulation was assessed via z-score distribution relative to sex-matched controls. We conducted single-subject KEGG pathway and AD biodomain functional enrichment analysis and utilized an unsupervised graph-based clustering approach to stratify patients into subgroups based on similarity of pathway alterations. Correlations between clinical metrics and omics signatures were evaluated at individual and subgroup levels.

Our single-subject approach revealed individualized, sex-dependent functional dysregulation in AD. In the transcriptomic profile, the peroxisome pathway was primarily downregulated in males (15% males vs. 3% females), while the MAPK signaling pathway was upregulated in females (17% females vs. 3% males). Endolysosome-related processes were predominantly upregulated in male AD individuals, whereas myelination-related subdomains were more pronounced in females. Apoptosis and epigenetic-related biodomains distinguished AD from asymAD, suggesting links to disease progression. Clustering analyses stratified individuals into molecular subtypes defined by cell cycle, DNA repair, synapse, and endolysosome-related subdomains, capturing heterogeneity beyond traditional diagnoses. Amyloid burden is correlated with oxidative stress in AD but not asymAD. Proteomics analyses aligned with transcriptomics findings, highlighting epigenetic regulation; and additional proteomics-specific modules indicative of post-transcriptional mechanisms.

Our findings underscore the molecular heterogeneity of AD, revealing individual and sex-dependent functional dysregulation better captured through single-subject analyses than cohort-based approaches. Clustering identified biologically distinct subgroups, some correlating with cognitive measures, providing potential biomarkers for refined disease stratification. These insights offer a framework for precision medicine in AD, guiding targeted therapeutic strategies based on individualized omics molecular signatures.