Analysis and validation of prognostic markers for thyroid cancer by single-cell and bulk RNA sequencing

Thyroid cancer represents the most prevalent malignancy within the endocrine system, exhibiting consistent increasing incidence rates over the years. We aim to utilize single-cell RNA sequencing (scRNA-seq) dataset to analyze key genes associated with prognosis in thyroid cancer.

Differential expression analysis and weighted gene co-expression network analysis (WGCNA) were applied to detect thyroid cell-related genes in the paratumor and tumor groups. Least absolute shrinkage and selection operator (LASSO) analysis was conducted on The Cancer Genome Atlas (TCGA) data to identify key genes and develop prognostic models. Immune infiltration analysis, correlation analysis, and differential expression analysis were used to explore the immunological features, correlation, and expression levels of the characterized genes. Western blot and immunohistochemical staining experiments were used to further validate the expression patterns of the characterized genes.

A total of 3,681 differentially expressed genes (DEGs) were recognized. There were 157 genes derived from the key modules in the WGCNA analysis. By performing intersection and conducting LASSO analysis, a prognostic model was developed based on the expression levels of six signature genes (APOE, MDK, NPC2, FCGBP, GLUL, and DTX4). The model's applicability was confirmed using the Gene Expression Omnibus (GEO) validation set. Immune infiltration analysis revealed varying degrees of immune cell infiltration between the high- and low-risk groups, which were intimately associated with the characterized genes. Subsequent experiments confirmed that APOE, MDK, and NPC2 protein levels were significantly higher in the thyroid cancer group compared to the control group.

APOE, MDK, NPC2, FCGBP, GLUL, and DTX4 may be key genes in thyroid cancer progression and serve as potential targets for disease treatment.