Development and experimental validation of hypoxia-related gene signatures for osteosarcoma diagnosis and prognosis based on WGCNA and machine learning

Q: Which methods were used to identify hypoxia-related diagnostic biomarkers for osteosarcoma?

The study used Gene Set Enrichment Analysis (GSEA), Weighted correlation network analysis (WGCNA), and LASSO, followed by ROC, decision curve analysis (DCA), and calibration curves to evaluate diagnostic performance.

Q: What immune-cell finding links STC2 to osteosarcoma risk?

Immune infiltration analysis showed cancer-associated fibroblasts (CAFs) were significantly higher in the low-risk group, and CAF infiltration was negatively associated with STC2 expression (P < 0.05).

Sep 21, 2026

Uploaded by Oliver

Osteosarcoma is a common malignant bone tumor with high mortality. This study comprehensively analyzes hypoxia signaling in osteosarcoma and develops hypoxia-related gene signatures to support prediction of diagnosis and prognosis. Gene Set Enrichment Analysis, WGCNA, and LASSO identify STC2 and TMEM45A as diagnostic biomarkers, evaluated using ROC, DCA, and calibration curves. Prognostic modeling relies on univariate and multivariate Cox regression, with risk scoring and survival assessment validated in external datasets and cell lines. Immune infiltration analysis links low-risk groups to higher CAFs.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages15

Reading Time38 min

AccessFree to download

Materials and methods

Data source

Identification and visualization

[www. nature.com/scientificreports](www. nature.com/scientificreports) OPEN Development and experimental validation of hypoxia‑related gene signatures for osteosarcoma diagnosis and prognosis based on WGCNA and machine learning BoWen1,2,4, Jian Chen1,4, Tianqi Ding1, Zhiyou Mao2, Rong Jin2, Yirui Wang3, Meiqin Shi2, Lixun Zhao2, AsangYang2, Xianyun Qin2* & Xuewei Chen1* Osteosarcoma (OS) is the most common primary malignant tumour of the bone with high mortality. Here, we comprehensively analysed the hypoxia signalling in OS and further constructed novel hypoxia‑related gene signatures for OS prediction and prognosis. This study employed Gene Set Enrichment Analysis (GSEA), Weighted correlation network analysis (WGCNA) and Least absolute shrinkage and selection operator (LASSO) analyses to identify Stanniocalcin 2 (STC2) and Transmembrane Protein 45A (TMEM45A) as the diagnostic biomarkers, which further assessed by Receiver Operating Characteristic (ROC), decision curve analysis (DCA), and calibration curvesin training and test dataset. Univariate and multivariate Cox regression analyses were used to construct the prognostic model. STC2 and metastasis were devised to forgethe OS risk model. Thenomogram, risk score, Kaplan Meier plot, ROC, DCA, and calibration curves results certified the excellent performance of the prognostic model. The expression level ofSTC2 and TMEM45A was validated in external datasets and cell lines. In immune cell infiltration analysis, cancer‑associated fibroblasts (CAFs) were significantly higher in the low‑risk group. And the immune infiltration of CAFs was negatively associated with the expression of STC2 (P < 0.05) . Pan‑cancer analysis revealed that the expression level ofSTC2 was significantly higher in Esophageal carcinoma (ESCA), Head and Neck squamous cell carcinoma (HNSC), Kidney renal clear cell carcinoma (KIRC), Lung squamous cell carcinoma (LUSC), and Stomach adenocarcinoma (STAD). Additionally, the higher expression ofSTC2 was associated with the poor outcome in those cancers. In summary, this study identified STC2 and TMEM45A as novel markers for the diagnosis and prognosis of osteosarcoma, and STC2 was shown to correlate with immune infiltration of CAFs negatively. Keywords Osteosarcoma, Hypoxia, Diagnosis, Prognosis, WGCNA, LASSO Osteosarcoma (OS) is the most common primary malignant tumour of the bone, usually occurring at the knee and proximal humerus. OS incidence has a bimodal age distribution with peaks in the 10–19 and 60–79 age groups1. Although rare, the incidence of OS is reported to be 2–4.2 cases per million population per year, the clinical outcome of OS has always been poor, and a high degree of OS has high mortality2. Furthermore, the underlying mechanism and potential therapeutic targets remain to be fully elucidated. Hence, developing new predictive and prognostic biomarkers for OS is essential. It has been known for many years that hypoxia condition is the hallmark of most solid tumors3. Under hypoxia, hypoxia-inducible factors (HIF) become activated and further bind to hypoxia-response elements (HRE) in target genes4. Thus, hypoxia regulates tumour biological processes, including angiogenesis, cell survival, metabolism and metastasis5. TMEM45A is a member of the transmembrane protein family, with 275 1Tianjin Institute of Environmental and Operational Medicine, Tianjin 300050, China. 2Department of Orthopedics, No. 945 Hospital of the PLA Joint Logistics Support Force, Yaan 625000, Sichuan, China. 3Department of Cardiology, No. 945 Hospital of the PLA Joint Logistics Support Force, Yaan 625000, Sichuan, China. 4These authors contributed equally: BoWen and Jian Chen.* email: [20188739@qq.com](20188739@qq.com); [chenxuewei11@sina.com](chenxuewei11@sina.com) [www. nature.com/scientificreports/](www. nature.com/scientificreports/) amino acids, which is highly expressed in the skin and participates in epidermal keratinization6,7. TMEM45A has been report

FAQs

Which methods were used to identify hypoxia-related diagnostic biomarkers for osteosarcoma?

The study used Gene Set Enrichment Analysis (GSEA), Weighted correlation network analysis (WGCNA), and LASSO, followed by ROC, decision curve analysis (DCA), and calibration curves to evaluate diagnostic performance.

How was the prognostic model constructed?

Univariate and multivariate Cox regression analyses were used to construct the prognostic model, producing a risk score and nomogram, then assessed with Kaplan–Meier plots and related performance curves.

What immune-cell finding links STC2 to osteosarcoma risk?

Immune infiltration analysis showed cancer-associated fibroblasts (CAFs) were significantly higher in the low-risk group, and CAF infiltration was negatively associated with STC2 expression (P < 0.05).

Sep 21, 2026

Uploaded by Oliver