Diagnostic Model for Alzheimer's Disease Based on PANoptosis-Related Genes.

Abstract

The pathophysiology of Alzheimer's disease (AD) involves the interplay of three different processes: pyroptosis, apoptosis, and necroptosis. To explore role of PANoptosis, a novel pro-inflammatory programmed cell death pathway, in AD patients. We performed a consensus clustering analysis to identify distinct transcriptional profiles in the samples using the R package "ConsensusClusterPlus". The PANoptosis key genes were obtained by crossing the WGCNA brown module and differentially expressed PANoptosis genes. We accomplished regression analyses using the LASSO-Cox method, combined with pathological status and gene expression data. At the same time, we also constructed PANscore system. The expression of PANoptosis hub genes were validated by qRT-PCR in AD transgenic mice. Our study utilized tissue expression profile data from AD patients to construct three distinct PANoptosis patterns, each with unique molecular and clinical characteristics. We have created a risk scoring system called PANscore, which can analyze patterns specific for each AD patient. Additionally, we observed significantly lower levels of follicular helper T (Tfh) cells in the high PANscore and AD patients. Further analysis revealed a significant negative correlation of Tfh with GSDMD and MLKL. These findings provide a roadmap for personalized patient stratification, enabling clinicians to develop personalized treatment plans for AD patients and advance the field of precision medicine.