Identification of Lactylation-Related Diagnostic Biomarkers for Atherosclerosis via Integrative Transcriptomic and Machine Learning Analysis

Introduction: Atherosclerosis (AS) is the primary pathological driver of cardiovascular disease mortality. Lactylation modification plays a critical role in cardiovascular diseases, the potential of lactylation-related genes (LRGs) as novel diagnostic biomarkers for AS remains to be fully elucidated. Objectives This study aims to identify key LRGs for the diagnosis of AS. Methods We analyzed the transcriptome dataset of AS patients from the GEO database, identified key genes and co-expression modules through differential expression analysis and Weighted Gene Co-Expression Network Analysis (WGCNA), and selected key LRGs related to AS by integrating the MSigDB database. Establish a foam cell model to verify these key LRGs. Machine learning diagnostic models (random forest, support vector machine, and generalized linear model) were constructed to evaluate the importance of key LRGs in AS. CIBERSORT and xCell were used to assess the cell composition of AS patients, with a focus on LRGs related to macrophage function. Results Our integrated analysis identified four key LRGs (KCNN4, PIK3CG, SLC25A4, and TPP1) that were significantly dysregulated in AS, and their expression was successfully validated in the macrophage model. The machine learning models further confirmed the importance of these four LRGs. Immune infiltration analysis revealed that TPP1 may be a key gene involved in macrophage function within the AS context. Conclusion This study identified and verified that KCNN4, PIK3CG, SLC25A4, and TPP1 are key LRGs related to AS, highlighting the potential of these LRGs as diagnostic biomarkers and therapeutic targets for AS.