Identification and validation of potential common biomarkers for papillary thyroid carcinoma and Hashimoto's thyroiditis through bioinformatics analysis and machine learning

Abstract

There is a growing body of evidence suggesting that Hashimoto's thyroiditis (HT) may contribute to an increased risk of papillary thyroid carcinoma (PTC). However, the exact relationship between HT and PTC is still not fully understood. The objective of this study was to identify potential common biomarkers that may be associated with both PTC and HT. Three microarray datasets from the GEO database and RNA-seq dataset from TCGA database were collected to identify shared differentially expressed genes (DEGs) between HT and PTC. A total of 101 genes was identified as common DEGs, primarily enriched inflammation- and immune-related pathways through GO and KEGG analysis. We performed protein-protein interaction analysis and identified six significant modules comprising a total of 29 genes. Subsequently, tree hub genes (CD53, FCER1G, TYROBP) were selected using random forest (RF) algorithms for the development of three diagnostic models. The artificial neural network (ANN) model demonstrates superior performance. Notably, CD53 exerted the greatest influence on the ANN model output. We analyzed the protein expressions of the three genes using the Human Protein Atlas database. Moreover, we observed various dysregulated immune cells that were significantly associated with the hub genes through immune infiltration analysis. Immunofluorescence staining confirmed the differential expression of CD53, FCER1G, and TYROBP, as well as the results of immune infiltration analysis. Lastly, we hypothesise that benzylpenicilloyl polylysine and aspirinmay be effective in the treatment of HT and PTC and may prevent HT carcinogenesis. This study indicates that CD53, FCER1G, and TYROBP play a role in the development of HT and PTC, and may contribute to the progression of HT to PTC. These hub genes could potentially serve as diagnostic markers and therapeutic targets for PTC and HT.

Figure 1

Flow chart of the whole study.

Figure 2

Differential expression gene analysis, function enrichment analysis and pathway enrichment analysis. (a) The PCA plot of GSE35570. (b, c) The Volcano plot and heatmap of DEGs in GSE33570. (d) The PCA plot of GSE29315. (e, f) The Volcano plot and heatmap of DEGs in GSE29315. (g) Venn plot of the up-regulated DEGs. (h) Venn plot of the down-regulated DEGs. (i) The KEGG enrichment analyses of DEGs. (j) The GO enrichment analyses of DEGs.

Figure 3

Screening of hub genes and the diagnostic value of hub genes. (a) The rankings of gene importance in GSE35570. (b) The rankings of gene importance in GSE29315. (c) Venn plot of the top eight genes in GSE35570 and GSE29315. (d) Diagnostic value of hub genes in the GSE35570. (e) Diagnostic value of hub genes in the GSE29315, (f) Diagnostic value of hub genes in the TCGA. (g) Diagnostic value of hub genes in the GSE138198.

Figure 4

ANN model construction and feature importance analysis. (a) The ANN was constructed based on the shared hub genes. (b) Diagnostic value of the ANN model in the GSE35570. (c) Diagnostic value of the ANN model in the TCGA. (d) Diagnostic value of the ANN model in the GSE138198. (e) A score calculated by SHAP was used for each input feature. (f, g) Distribution of the impact of each feature on the full model output estimated using the SHAP values.

Figure 5

Microscopy scan of IF staining showed the distribution of CD53(green), FCER1G(green), and TYROBP(green), in HT-related PTC tissues and normal tissues adjacent to the tumour (NAT); as well as diagnostic value of CD53, FCER1G and TYROBP. MFI: Mean Fluorescence Intensity.

Figure 6

Microscopy scan of IF staining showed the distribution of Cd4(green), Cd8(green), and Cd86(green), in HT-related PTC tissues and normal tissues adjacent to the tumour (NAT). MFI: Mean Fluorescence Intensity.