Identification of Mitochondrial-related Characteristic Biomarkers inOsteosarcoma using Bioinformatics and Machine Learning

Le résumé fourni par la source

Background/Aims: Osteosarcoma (OS), a malignant tumor originating in bone or cartilage, primarily affects children and adolescents. Notably, substantial alterations in mitochondrial energy metabolism have been observed in OS; however, the specific contribution of mitochondrial- related genes (MRGs) to OS pathogenesis and prognosis remains unclear. Herein, we identified novel diagnostic biomarkers associated with mitochondrial-related processes in OS via comprehensive bioinformatics analysis. Methods: OS mRNA expression profiles were retrieved from GSE16088 and GSE19276 databases. Mitochondrial-related differentially expressed genes (MitoDEGs) were identified by integrating differentially expressed analysis with mitochondrial-localized genes. A protein-protein interaction network was constructed, and machine learning algorithms (LASSO regression analysis and SVM-RFE) identified characteristic MitoDEGs. Subsequently, immune cell infiltration, microenvironment analysis, and single-cell RNA sequencing (scRNA-seq) analyzed differences in characteristic MitoDEGs, and RT-PCR was used for in vitro verification of characteristic MitoDEGs. Results: MitoDEGs in OS were significantly enriched in the pathways associated with mitochondrial function and immune regulation. Two MitoDEGs, UCP2 and PRDX4, were identified via LASSO and SVM-RFE. Correlation analysis demonstrated a close association between UCP2 and PRDX4 expression levels and immune cell infiltration, particularly in CD8+ T and native CD4+ T cells, as observed in both immune cell and scRNA-seq analyses. Furthermore, RTPCR confirmed the expression levels of UCP and PRDX4 at the cellular level, which was consistent with the bioinformatics results. Conclusion: This study identified UCP2 and PRDX4 as characteristic MitoDEGs and potential diagnostic biomarkers for OS using machine learning algorithms. These findings provide novel insights into the clinical applications of these biomarkers for OS diagnosis.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.