Additional file 1 of In silico analysis of differentially expressed genesets in metastatic breast cancer identifies potential prognostic biomarkers

Additional file 1: Supplementary Figure S1. KIF2C and ESR1 are hub genes the most significantly co-expressed with the potential biomarker candidate genes. (A and B) Shown are Pearson's pairwise correlation plots of RNA-seq gene expression between four DOE-A (A) or three DUE-A genes (B) and their most significantly co-expressed hub genes identified from PPI networks. Statistical analyses were performed by the pre-set analytic method of bc-GenExMiner. Supplementary Figure S2. Comparison of mRNA expression of the two most significantly co-expressed hub genes (KIF2C and ESR1) between basal-like or triple-negative breast cancer and other subtypes of breast cancer. (A and B) RNA-seq data of KIF2C and ESR1 were obtained from the Cancer Cell Line Encyclopedia (CCLE) and analyzed. N = 31 in BL/TNBC and N = 26 in luminal type cell lines. (C and D) RNA-seq data of KIF2C and ESR1 from The Cancer Genome Atlas (TCGA) [58] were analyzed at bc-GenExMiner v4.3. N = 97 in BL/TNBC and N = 736 in non-BL/TNBC type breast cancer patient samples. Statistical significance in A and B was determined by unpaired t-tests and those in C and D were determined by the pre-set analytic method of bc-GenExMiner. Supplementary Figure S3. Correlation between the expression levels of two co-expressed hub genes (KIF2C and ESR1) and patient survivals. (A and B) Relapse-free, overall, distant metastasis-free, and post-progression survival of two co-expressed hub genes (KIF2C in (A); ESR1 in (B)) were stratified by the expression levels of each gene (low or high). Expression data were analyzed by KM plotter ( http://kmplot.com/ ). JetSet best probes were selected and patients (for KIF2C, N = 3951 in RFS, = 1402 in OS, = 1746 in DMFS and = 414 in PPS; for ESR1, N = 3951 in RFS, = 1402 in OS, = 1746 in DMFS and = 414 in PPS) were split by median expression. (C) Metastatic relapse-free survival of KIF2C and ESR1 was stratified by the expression levels of each gene (low or high). Microarray expression data were analyzed by bc-GenExMiner v4.3 ( http://bcgenex.centregauducheau.fr/ ). Patients (KIF2C, N = 4533; ESR1, N = 4785) were split by median expression. Statistical analyses were performed by pre-set analytic methods. HRs (hazardous ratios) and 95% CIs (confidence intervals) are indicated.

附加文件1：补充图S1。KIF2C与ESR1为与潜在候选生物标志物基因存在最显著共表达关系的枢纽基因（hub gene）。(A、B) 本图展示了4个DOE-A基因（A）或3个DUE-A基因（B）的RNA测序（RNA-seq）基因表达量，与其从蛋白质相互作用（Protein-Protein Interaction, PPI）网络中鉴定出的最显著共表达枢纽基因之间的Pearson两两相关性散点图。统计分析采用bc-GenExMiner预设的分析方法完成。

补充图S2：对比基底样型或三阴性乳腺癌（basal-like or triple-negative breast cancer, BL/TNBC）与其他乳腺癌亚型中，两个最显著共表达的枢纽基因（KIF2C与ESR1）的mRNA表达水平。(A、B) KIF2C与ESR1的RNA测序数据取自癌症细胞系百科全书（Cancer Cell Line Encyclopedia, CCLE）并进行分析：基底样/三阴性乳腺癌细胞系样本量N=31，腔面型细胞系样本量N=26。(C、D) 来自癌症基因组图谱（The Cancer Genome Atlas, TCGA）[58]的KIF2C与ESR1 RNA测序数据在bc-GenExMiner v4.3中完成分析：基底样/三阴性乳腺癌患者样本量N=97，非基底样/非三阴性乳腺癌患者样本量N=736。A、B组的统计学显著性采用非配对t检验确定，C、D组的统计学显著性采用bc-GenExMiner预设的分析方法确定。

补充图S3：两个共表达枢纽基因（KIF2C与ESR1）的表达水平与患者生存期的相关性分析。(A、B) 根据每个基因的表达水平（高表达组与低表达组）进行分层，展示两个共表达枢纽基因（A为KIF2C，B为ESR1）的无复发生存期（Relapse-Free Survival, RFS）、总生存期（Overall Survival, OS）、远处无转移生存期（Distant Metastasis-Free Survival, DMFS）及进展后生存期（Post-Progression Survival, PPS）。表达数据通过KM绘图工具（KM plotter，http://kmplot.com/）进行分析。选取JetSet最优探针，患者样本量如下：KIF2C组（RFS：N=3951，OS：N=1402，DMFS：N=1746，PPS：N=414）；ESR1组（RFS：N=3951，OS：N=1402，DMFS：N=1746，PPS：N=414），所有患者按基因表达中位数进行分组。(C) 根据每个基因的表达水平（高表达组与低表达组）进行分层，展示KIF2C与ESR1的转移性无复发生存期。芯片表达数据通过bc-GenExMiner v4.3（http://bcgenex.centregauducheau.fr/）进行分析。患者样本量：KIF2C组N=4533，ESR1组N=4785，所有患者按基因表达中位数进行分组。统计分析采用预设分析方法完成。图中标注了风险比（hazard ratios, HRs）与95%置信区间（confidence intervals, CIs）。