Identification of potential molecular targets for thyroid cancer using multiple-omics analysis and machine learning

Thyroid cancer is the most prevalent endocrine tumor, with increasing incidence rates annually. Identifying novel molecular markers is crucial for early diagnosis and personalized treatment.

Differentially expressed genes (DEGs) from thyroid cancer microarray data were analyzed, functional enrichment analysis was performed, and machine learning was used to identify signature genes. Weighted gene co-expression network analysis (WGCNA) was applied to determine key modules associated with thyroid cancer. The expression and correlation of identified genes were analyzed, and their immune infiltration and clinical characteristics were assessed. Single-cell RNA transcriptomics was used to analyze the expression patterns of genes. The expression levels of candidate genes in thyroid cancer specimens were verified by immunohistochemistry and RT-PCR.

A total of 348 DEGs were identified, with 18 feature genes selected by machine learning. WGCNA revealed 12 key genes associated with thyroid cancer. Among these, ERO1B, FN1, and CRABP1 were significantly linked to clinical features, patient survival, and tumor immune infiltration. Single-cell RNA transcriptomic analysis showed distinct expression patterns for these genes across different cell types. To verify our findings, the immunohistochemical and RT-PCR results showed that the expression of CRABP1 was significantly downregulated, and the expression of FN1 was significantly upregulated in thyroid cancer tissues.

FN1 and CRABP1 were identified as potential molecular targets in thyroid cancer. These genes influence tumor progression, immune response, and patient prognosis, offering new insights for early diagnosis and personalized treatment strategies.