Fig 1.
Expression and mutation status of HuR were evaluated in normal tissue, cancer cell lines, and pan-cancer samples.
(A) Expression of HuR in normal human tissues, including Thyroid, Ovary, and Lung et al, were performed with GTEx dataset. (B) Expression of HuR in pan-cancer tissues, including 33 human cancer types accessed from TCGA dataset, were displayed with boxplots. (C) Expression of HuR in pan-cancer tissues were accessed in paired cancer tissues. (D) Expression of HuR in pan-cancer tissues were detected in GEO dataset. (E) Mutation status of HuR across pan-cancer were displayed with bar graph. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
Fig 2.
Relationship between HuR and patients’ outcome was estimated with pan-cancer overall survival (OS) analysis.
Log-rank test OS analysis of HuR was displayed with survival curve in different cancer types, including ACC (A), BRCA (B), GBM-LGG (C), LGG (D), LIHC (E), LUAD (F), MESO (G), PAAD (H), SARC (I), and SKCM (J).
Fig 3.
The diagnostic value of HuR across pan-cancers was accessed with ROC analysis and TMB analysis.
(A) The ROC curves of HuR across cancers were evaluated in BLCA, BRCA, CESC, COAD, ESCA, HNSC, KICH, LIHC, LUAD, LUSC, PCPG, PRAD, READ, SARC, STAD, and UCEC. AUC: area under curve. (B) Pearson correlation analysis between HuR and TMB across cancers was performed with scatter diagram. (C) Landscape of pearson correlation between HuR and TMB was displayed with radar plot. R, Pearson correlation coefficient. P, P-value of Pearson correlation. TMB, tumor mutation burden; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
Fig 4.
Correlation between HuR and PD-L1 was estimated in normal and tumor tissues, separately.
Correlation between HuR and PD-L1 in normal adrenal gland and ACC (A), breast tissue and BRCA (B), colon tissues and COAD (C), esophageal and ESCA (D), normal kidney and KICH (E), normal brain tissue, GBM, and LGG (F), normal lung tissue, LUAD, and LUSC (G), normal skin and HNSC (H), normal pancreas and PAAD (I), normal stomach tissue and STAD (J), normal thyroid tissue and THCA (K). R, Pearson correlation coefficient. P, P-value of Pearson correlation.
Fig 5.
Effect of HuR on tumor immune cells across pan-cancer were estimated with Immune Score and ssGSEA immune Infiltration analysis.
(A) The variation of immune scores in different groups with high and low expression of HuR were displayed with violin plots. (B) ssGSEA immune infiltration analysis of HuR was accessed in different cancer types and showed in heatmap. Immune Score, an indicator for assessing the degree of immune cell infiltration in tumors was calculated with Estimate package. (B) Correlation between HuR and TIDE (Tumor Immune Dysfunction and Exclusion) score was displayed with boxplots across pan-cancer. *p < 0.05, **p < 0.01, ***p < 0.001.
Fig 6.
Effect of HuR on cancer drugs sensitivity across cancers was displayed with correlation circular clustering heatmap.
Spearman correlation analysis was used to estimate the correlation between HuR and sensitivity of different cancer drugs.
Fig 7.
Pan-cancer functional analysis of HuR was performed.
(A) Grouped volcano plot was used to displayed the alterations of gene expression profile between high or low level of HuR. (B) GSEA of HuR related DEGs (FC ≥ 1.5, p ≤ 0.05) was performed with bubble plot.