The rapid increase in the incidence of Alzheimer’s disease (AD) has raised concerns, given its profound effects on both society and the economy. Despite extensive research efforts in this area, there are no existing treatments that have the ability to change the progression of the disease.

To identify the distinct subtypes of AD, consensus clustering was employed. Following this, module genes were identified through the implementation of WGCNA. In addition, the investigation included the identification of hub genes through the application of machine learning. Ultimately, a thorough analysis was performed utilizing a methodical strategy to perform post-translational modification (PTM) genome wide.

GO and KEGG analyses were conducted by examining of 21 different types of PTMs, revealing that the majority of these genes play key roles in the PTM pathways, as well as AD-related pathways. Correlation analysis revealed that these PTM were significantly correlated with gamma secretase activity, beta secretase activity, amyloid-beta 42, clinical dementia rating, Braak stage, plaque, and neurofibrillary tangle. Then, two distinct subtypes of PTM were identified, each characterized by unique clinical characteristic. By utilizing machine learning, we developed an PTM.score, and has shown impressive predictive capabilities for AD when tested against various datasets (brain AUC: 0.859, blood AUC: 0.898), indicating its potential utility in clinical settings for risk stratification and therapeutic decision-making. Moreover, our investigation led to the identification of two genes (TRIM47 and LNX1) that may represent potential drug targets for AD (brain tissues or blood samples). Research further indicated a potential correlation between TRIM47 and concentrations of CSF Aβ (OR 1.068 (1.029–1.108)), CSF p-tau (OR 1.315 (1.136–1.524)), and total hippocampal (OR 1.176 (1.058–1.307)).

The findings from this study not only enhance our comprehension of the underlying mechanisms of AD but also serve to inform and direct future initiatives in drug discovery. By focusing on TRIM47, the work paves the way for identifying innovative therapeutic strategies.