Role of fatty acid metabolism-related genes in periodontitis based on machine learning and bioinformatics analysis.

Abstract

Objectives: This study aims to investigate the role of genes related to fatty acid metabolism in periodontitis through machine learning and bioinformatics methods.

Methods: Periodontitis datasets GSE10334 and GSE-16134 were downloaded from the GEO database, and the fatty acid metabolism-related gene sets were obtained from the GeneCards database. Differentially expressed fatty acid metabolism-related genes (DEFAMRGs) in periodontitis were screened using the "limma" R package. Functional enrichment and pathway analyses were conducted. Recursive Feature Elimination, Least Absolute Shrinkage and Selection Operator, and Boruta algorithm were used to determine hub DEFAMRGs and construct diagnostic models with internal and external validation. Subtypes of periodontitis related to hub DEFAMRGs were constructed using consistency clustering analysis. CIBERSORT was used to analyze immune cell infiltration in gingival tissues and explore the correlation between hub DEFAMRGs and immune cells.

Results: A total of 113 periodontitis DEFAMRGs were screened out as a result. The enrichment analysis results indicate that DEFAMRGs are mainly associated with immune inflammatory responses and immune cell chemotaxis.Finally, 8 hub DEFAMRGs (BTG2, CXCL12, FABP4, CLDN10, PPBP, RGS1, LGALSL, and RIF1) were identified and a diagnostic model (AUC=0.967) was constructed, based on which periodontitis was divided into two subtypes. In addition, there is a significant correlation between hub DEFAMRGs and different immune cell populations, with mast cells and dendritic cells showing higher correlation.

Conclusions: This study provides new insights and ideas for the occurrence and development mechanism of periodontitis and proposes a diagnostic model based on hub DEFAMRGs to provide new directions for diagnosis and treatment.