Transcriptome combined with single-cell data to construct a prognostic model for glycosylation-related genes in...

Area of Science:

• Oncology
• Genetics
• Bioinformatics

Background:

• Osteosarcoma (OS) is a primary bone cancer originating from mesenchymal cells.
• Glycosylation plays a crucial role in cancer cell adhesion, metastasis, and transformation.
• Research on glycosylation-related genes (GRGs) in OS remains limited.

Purpose of the Study:

• To identify prognostic glycosylation-related genes (GRGs) in osteosarcoma (OS).
• To develop a prognostic model for OS based on identified GRGs.
• To explore the molecular mechanisms and immune microenvironment associated with OS prognosis.

Main Methods:

• Utilized OS datasets (GSE36001, TARGET-OS, GSE21257) and a scRNA-seq dataset (GSE152048) with 582 GRGs.
• Employed differential expression analysis and Weighted Gene Co-expression Network Analysis (WGCNA) for gene identification.
• Applied univariate Cox regression and LASSO regression for prognostic gene filtering and risk score generation.

Main Results:

• Identified seven GRGs (DCN, RENBP, UAP1, B4GALNT1, CCDC115, TUBA1A, B3GALT4) as prognostic markers for OS.
• Developed a risk model categorizing patients into high- and low-risk groups, with the low-risk group showing improved survival.
• Found activated immune and apoptosis pathways in the high-risk group, correlating with poorer prognosis; identified osteoblastic cells as key cells involved in cell-cell communication.

Conclusions:

• Successfully identified seven prognostic GRGs for osteosarcoma (OS).
• Constructed prognostic models and explored underlying molecular mechanisms.
• These findings provide a valuable benchmark for OS prognosis and treatment strategies.