Gene Expression Profiling Interactive Analysis, Version 2 (GEPIA2) (web: http://gepia2.cancer-pku.cn/#analysis) (Tang Z. et al., 2019)

For certain tumors without correlated normal tissues (e.g., TCGA-DLBC, TCGA-UCS) in the TCGA project, we obtained the corresponding normal tissues in the Genotype-Tissue Expression (GTEx) database and used the “Expression Analysis-Box Plots” module to obtain box plots of the expression differences between these tumor tissues and the normal tissues (settings: P-value Cutoff = 0.01, Ilog2FCl Cutoff = 1, and “Match TCGA normal and GTEx data”).

We obtained violin plots of CBX8 expression in different pathological stages (stage I, stage II, stage III, and stage IV) of all TCGA tumors with the “Pathological Stage Plot” module of GEPIA2. The log2 [transcripts per million (TPM) + 1] transformed expression data were applied for the box or violin plots.

The “Survival Analysis” module was used to evaluate the significant relationship between overall survival (OS) and disease-free survival (DFS) and CBX8 expression for all tumors in the TCGA project (settings: cutoff-high: 50%, cutoff-low: 50%). The hypothesis tests used log-rank tests. These data were visualized by survival plots.

In the TCGA project, we obtained the top 100 target genes correlated with CBX8 via the “Similar Gene Detection” module and used pairwise Pearson correlation analysis to analyze the correlation between CBX8 and these target genes. For the two highest correlation coefficients (R), we used a dot plot to visualize the data.