Decoding lactylation in neuropathic pain: Immune cell infiltration patterns and machine learning-identified candidate biomarkers

This study aimed to identify lactylation-associated genes linked to immune infiltration and diagnostic potential in neuropathic pain using integrated bioinformatic and machine learning approaches. Two microarray datasets (GSE124272 and GSE150408) comprising peripheral blood transcriptomes from 25 NP patients and 25 healthy controls were obtained from the gene expression omnibus. After batch correction and merging, the combined dataset served as the training set. Differentially expressed genes overlapping with lactylation-related gene sets were identified. Functional enrichment analyses, including gene ontology and Kyoto encyclopedia of genes and genomes pathway analyses, were performed. A protein-protein interaction network was constructed. Three machine learning algorithms-least absolute shrinkage and selection operator, support vector machine recursive feature elimination, and random forest-were applied to identify robust diagnostic gene signatures. Subsequently, the candidate biomarkers were validated using an independent test set (GSE95849). A diagnostic nomogram was developed, and regulatory networks were analyzed. Immune infiltration analysis was conducted via cell-type identification by estimating relative subsets of RNA transcripts. Functional analyses indicated involvement of pathways such as glucagon signaling, thermogenesis, and mitochondrial inner membrane function. Machine learning-identified 5 diagnostic gene candidates: CYP27A1, ELAC2, TMEM126B, LYRM7, and PHKB. Among these, CYP27A1 and PHKB were further investigated in an independent test set. Immune infiltration analysis showed significant alterations in 19 immune cell types, with CYP27A1 and PHKB closely correlated with immune cell distribution. This study identified CYP27A1 and PHKB as potential lactylation-associated biomarkers for NP, offering new insights into its pathogenesis and a theoretical basis for improved diagnosis.