Fig 2: Prognostic Analysis and Survival Prediction of ABCC3 in Glioma. KM survival curves with corresponding confidence intervals illustrated the OS difference between ABCC3 high group (red curve) and ABCC3 low group (blue curve) in gliomas from TCGA (A), CGGA_693 (B) and CGGA_325 (C). Meta-analysis of ABCC3 expression and OS of glioma samples from TCGA, CGGA_693 and CGGA_325 with random effects model (D). Boxplots showing the expression of ABCC3 in various prognostic clinical features (Age, WHO Grade, IDH mutation status, 1p/19q co-deletion status and MGMT Methylated status) from TCGA (E), CGGA_693 (F) and CGGA_325 (G)

Fig 3: Functional Enrichment and Immune Correlation Analysis of ABCC3 in Glioma. The intersect genes of three differentially expressed genes sets from TCGA, CGGA_693 and CGGA_325 (A). Go function enrichment analysis of the intersect DEGs (B). Immune cells infiltration in ABCC3 low group (blue) and ABCC3 high group (red) in TCGA database (C). Lollipop chart displayed the correlation coefficient between ABCC3 expression and immune cells infiltration in the TCGA database (D). Correlation heat map of ABCC3 and 20 common immune checkpoint molecules. The red circles represented a positive correlation; the blue circles represented a negative correlation. The coefficients were bolded if the values were large than 0.5 (E)

Fig 4: Expression Analysis of ABCC3 in Glioma. Structural model of ABCC3 (A). Expression levels of ABCC3 RNA in WHO 2 (507 cases), WHO 3 (571 cases), and WHO 4 (540 cases) show statistically significant differences. Higher WHO grades correlate with increased levels of ABCC3 RNA (B). Representative sites (red rectangle) of tissue sampling for MR imaging in peripheral tissue, LGG, and Recurrent GBM, along with H&E staining and IHC for ABCC3 (white arrow) (C). ABCC3 protein expression was higher in GBM (n = 10) compared to LGG (n = 8) (D). (***p < 0.001)

Fig 5: Identification and Functional Analysis of ABCC3-Related Gene Subtypes in Glioma. Venn diagram displaying the 26 intersecting genes within ABCC3-related gene sets. The ABCC3-related genes were identified as MM > 0.8 and GS > 0.5 in the most ABCC3 high expression related module (TCGA_magenta, CGGA_693_pink, CGGA_325_blue) from three databases basing on WGCNA (A). The consensus map of the non-negative matrix factorization (NMF) for ABCC3 high group from TCGA identified two molecular subtypes (C1 and C2) according to 26 ABCC3-related genes (B). The KM survival curve accompanied by number at risk and cumulative number of events revealed survival rates in C1, C2, and C3(ABCC3 low group) (C). Go function enrichment analysis of the DEGs between C1 and C2 (D)