Identification and Characterization of Metastasis-Associated Individualized Gene Expression Signature in Osteosarcoma

Abstract

Background: Osteosarcoma (OS) patients with surgical resection still relapse with poor prognosis due to the inability to detect distant metastasis. It’s essential to identify metastasis-related biomarkers for OS. Methods: Here, a computational pipeline follow relative expression orderings (REOs) using gene expression was constructed in metastases and non-metastases OS patients. Results: 138 metastasis-associated gene pair signatures (MGPSs) were identified follow two independent datasets. A metastases-specific co-expressed MGPS network was constructed for extracting biomarker for clinical application. MGPS such as MYL5 and RPL27A showed strong positive correlation (Cor =0.75, P <0.001). There were thirteen prognostic MGPSs in above network. These prognostic MGPSs could become as a specific classifier to distinguish metastases and non-metastases OS patients. MGPSs were associated with cancer metastasis-related functions. Drug and MGPS network could provide some drug candidates for treatment of OS. Conclusions: Collectively, the roles of the MGPSs in OS were elucidated, which could be beneficial for understanding OS pathogenesis and treatment.