Exploratory bioinformatics analysis for identifying candidate biomarkers of type 1 diabetes mellitus

Abstract

Background Type 1 diabetes mellitus (T1DM) is a chronic disease that significantly impacts patients’ quality of life. Its prevalence is rising globally each year. This study aims to identify potential biomarkers associated with T1DM through comprehensive bioinformatics analysis, further enhancing T1DM early diagnosis and treatment. Methods Transcriptome datasets from T1DM patients and the control group were from the Gene Expression Omnibus (GEO) database. Differentially Expressed Genes (DEGs) were identified and subsequently analyzed using Gene Ontology (GO) enrichment, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment, and protein–protein interaction (PPI) network analysis. Hub genes were identified using Enzyme-Linked Immunosorbent Assay (ELISA) on clinical samples comprising 17 T1DM patients and 19 controls. Immune cell infiltration was estimated using the Cell-type Identification By Estimating Relative Subsets Of RNA Transcripts (CIBERSORT) algorithm, while the diagnostic performance of the hub genes was evaluated via receiver operating characteristic (ROC) curve analysis. Results A total of 20 up-regulated and eight down-regulated DEGs were identified in the GEO database. Functional enrichment analysis showed that immune activation played an important role in T1DM. The expression levels of the hub genes, CTSG and LTF, were further validated in clinical samples. ROC analysis showed moderate diagnostic performance, with AUC values of 0.75 (training set) and 0.67 (validation set). Conclusions The results indicate that CTSG and LTF may serve as promising diagnostic biomarkers for T1DM. Our study is positioned as exploratory with moderate diagnostic relevance rather than definitive biomarker discovery. The findings are preliminary and require further validation before any clinical application.