Supplementary Files

Table S1: Datasets used in this study. Table S2: Primer sequences used in this study. Table S3: 125 exosome-related genes from literature and databases. Table S4: 4355 OS-related lncRNAs by univariate cox regression analysis. Table S5: Genes of 15 exosome-related modules defined by a correlation coefficient over 0.3 in module-trait correlation analysis. Table S6: 42 hub genes of exosome-related modules defined by gene significance over 0.3 and module membership over 0.8. Table S7: Restricted mean survival time between risk groups at different time points. Table S8: Correlation between 26 modeling lncRNAs and 812 FDA approved drug targets. A correlation coefficient over 0.3 is considered exosome-related. Table S9: Correlation between 26 modeling lncRNAs and 1572 potential drug targets. A correlation coefficient over 0.3 is considered exosome-related. Table S10: Trend tests for survival analysis in different significance test methods. Table S11: Correlation between 26 modeling lncRNAs and 1055 OS-related RBPs. A correlation coefficient over 0.2 is considered exosome-related.

补充表S1：本研究使用的数据集。
补充表S2：本研究使用的引物序列。
补充表S3：来自文献与数据库的125个外泌体相关基因。
补充表S4：通过单变量Cox回归分析得到的4355个骨肉瘤（osteosarcoma, OS）相关长链非编码RNA（long non-coding RNA, lncRNA）。
补充表S5：在模块-性状相关性分析中，相关系数大于0.3所定义的15个外泌体相关模块的基因集。
补充表S6：以基因显著性（gene significance）大于0.3且模块隶属度（module membership）大于0.8所定义的外泌体相关模块的42个核心基因（hub gene）。
补充表S7：不同时间点风险组间的限制性平均生存时间（restricted mean survival time）。
补充表S8：26个建模用长链非编码RNA与812个经美国食品药品监督管理局（Food and Drug Administration, FDA）批准的药物靶点之间的相关性分析结果，相关系数大于0.3者视为外泌体相关。
补充表S9：26个建模用长链非编码RNA与1572个潜在药物靶点之间的相关性分析结果，相关系数大于0.3者视为外泌体相关。
补充表S10：不同显著性检验方法下的生存分析趋势检验。
补充表S11：26个建模用长链非编码RNA与1055个骨肉瘤相关RNA结合蛋白（RNA-binding protein, RBP）之间的相关性分析结果，相关系数大于0.2者视为外泌体相关。