Objective: To screen the prognostic biomarkers of metabolic genes in patients with multiple myeloma (MM), and construct a prognostic model of metabolic genes.
Methods: The histological database related to MM patients was searched. Data from MM patients and healthy controls with complete clinical information were selected for analysis.The second generation sequencing data and clinical information of bone marrow tissue of MM patients and healthy controls were collected from human protein atlas (HPA) and multiple myeloma research foundation (MMRF) databases. The gene set of metabolism-related pathways was extracted from Molecular Signatures Database (MSigDB) by Perl language. The biomarkers related to MM metabolism were screened by difference analysis, univariate Cox risk regression analysis and LASSO regression analysis, and the risk prognostic model and Nomogram were constructed. Risk curve and survival curve were used to verify the grouping effect of the model. Gene set enrichment analysis (GSEA) was used to study the difference of biological pathway enrichment between high risk group and low risk group. Multivariate Cox risk regression analysis was used to verify the independent prognostic ability of risk score.
Results: A total of 8 mRNAs which were significantly related to the survival and prognosis of MM patients were obtained (P<0.01). As molecular markers, MM patients could be divided into high-risk group and low-risk group. Survival curve and risk curve showed that the overall survival time of patients in the low-risk group was significantly better than that in the high risk group (P<0.001). GSEA results showed that signal pathways related to basic metabolism, cell differentiation and cell cycle were significantly enriched in the high-risk group, while ribosome and N polysaccharide biosynthesis signaling pathway were more enriched in the low-risk group. Multivariate Cox regression analysis showed that the risk score composed of the eight metabolism-related genes could be used as an independent risk factor for the prognosis of MM patients, and receiver operating characteristic curve (ROC) showed that the molecular signatures of metabolism-related genes had the best predictive effect.
Conclusion: Metabolism-related pathways play an important role in the pathogenesis and prognosis of patients with MM. The clinical significance of the risk assessment model for patients with MM constructed based on eight metabolism-related core genes needs to be confirmed by further clinical studies.
题目: 基于代谢相关基因的多发性骨髓瘤预后模型的构建.
目的: 筛选多发性骨髓瘤（MM）患者代谢相关基因预后生物标志物，构建MM患者代谢基因生存预后模型.
方法: 检索MM患者相关的组学数据库。选择具有完整临床信息的病例和健康对照组数据进行分析。从HPA与MMRF数据库收集整理MM患者与健康对照骨髓组织二代测序数据与临床信息。利用Perl语言从分子签名数据库（MSigDB）提取代谢相关通路基因集。利用差异分析、单因素Cox风险回归分析和LASSO回归分析筛选MM代谢相关预后生物标志物并构建风险预后模型及列线图，利用风险曲线与生存曲线验证模型分组效果。利用基因集富集分析（GSEA）研究高、低风险组之间生物学通路富集的差异。利用多因素Cox风险回归分析验证风险评分的独立预后预测能力.
结果: 共筛选获取8个与MM患者生存预后显著相关的mRNA（P<0.01），作为分子标签可将MM患者分为高风险组与低风险组。生存曲线与风险曲线显示低风险组患者的总生存期显著优于高风险组（P<0.001）。GSEA富集分析表明，基础代谢相关通路、细胞分化和细胞周期等信号通路在高风险组中显著富集，核糖体与N-聚糖生物合成等相关通路则更多的在低风险组中富集。多因素Cox回归分析显示，由这8个代谢相关基因构成的风险评分可作为MM患者预后的独立危险因素，接收者操作特征曲线显示代谢相关基因分子标签预测效能最佳.
结论: 新陈代谢相关通路在MM患者发病及预后中具有重要意义，基于8个代谢相关核心基因构建的MM患者风险评估模型，其临床意义尚需进一步的临床研究加以证实.
Keywords: gene set enrichment analysis; metabolism; multiple myeloma; risk prognostic model.