目的:建立基于乳腺癌脂肪酸代谢(fatty acid metabolism,FAM)相关基因的预后模型,并对乳腺癌的FAM 模型进行评估预后.方法:整理乳腺癌单细胞 RNA 测序(single-cell RNA sequencing,scRNA-seq)数据集,鉴定出乳腺上皮肿瘤细胞,通过表达水平曲线下面积(are-a under the curve for expression level,AUCell)分析得出乳腺癌FAM在不同肿瘤细胞内的活化程度,再与加权基因共表达分析(weighted gene co-expression network analysis,WGCNA)结果取交集得到乳腺癌FAM相关差异基因.通过单因素和多因素Cox回归建立 767 例患者的预后风险模型.多指标受试者工作特征曲线(receiver operating characteristic,ROC),决策分析曲线(deci-sion analysis curve,DAC),一致性指数(concordance index,C-index)用于评估模型的准确性.最后使用GSE3143 数据集验证.结果:单细胞组学分析发现597 个差异基因,单变量和多变量Cox回归确定了12 个与OS相关的风险相关基因:FOXQ1、TFPI2、MACC1、ACOT7、MCEE、SLC27A2、QPRT、SLC2A1、ACAA1、NDRG1、KYNU、YOD1.多指标ROC分析,DAC以及C-index均提示该模型预测准确率较高.结论:基于12 个乳腺癌FAM相关基因预后模型具有一定的准确性,可更好地用于指导临床治疗.
Establishing and verification of fatty acid metabolism prognosis model for breast cancer based on single-cell transcriptomic sequencing data
Objective:To establish a prognostic model of fatty acid metabolism(FAM)for breast cancer based on FAM related genes,and to evaluate the prognosis of the FAM model.Methods:Breast cancer single-cell RNA sequencing(scRNA-seq)data set was sorted out to identify breast epithelial tumor cells,and the activation degree of breast cancer FAM in different tumor cells was ob-tained using area under the curve for expression level(AUCell)analysis.The weighted gene co-ex-pression network analysis(WGCNA)was used to obtain FAM-related differential genes in breast cancer.The prognostic risk model of 767 patients was established using univariate and multivariate COX regression.Receiver operating characteristic curve(ROC),decision analysis curve(DAC)and concordance index(C-index)were used to evaluate the accuracy of the model.Finally,the GSE3143 dataset was used for validation.Results:A total of 597 differential genes were identified by single-cell transcriptomic analysis,and 12 risk-related genes associated with OS were identified by univariate and multivariate COX regression,which were FOXQ1,TFPI2,MACC1,ACOT7,MCEE,SLC27A2,QPRT,SLC2A1,ACAA1,NDRG1,KYNU and YOD1.Multi-index ROC analysis,DAC and C-index indicated that the model had high prediction accuracy.Conclusion:The prognostic model based on 12 FAM-related genes for breast cancer has certain accuracy,which could be used to guide clinical treatment more accurately.
万方数据
维普
