目的 整合胰腺导管腺癌（PDAC）分子亚型和免疫相关基因构建风险预测模型。方法以GSE71729数据集（n=145）为训练集，差异表达基因和差异免疫相关基因整合PDAC鳞状和非鳞状亚型之间的差异构建调控网络,在此基础上筛选出调控鳞状亚型的5个免疫标志基因,并根据患者生存信息构建综合免疫评分(IIS)模型和免疫标记基因来预测PDAC患者的临床预后，并用5个验证集（n=758）测试其预测性能。结果根据IIS将PDAC患者分为高风险组和低风险组。验证集显示，高危组患者总生存期短于低危组（均P<0.001）。多变量Cox回归显示IIS是PDAC的独立预后因素（HR=2.16，95% CI=1.50-3.10，P<0.001)。结论IIS可用于PDAC患者的危险分层，并可能成为PDAC潜在的预后标志物。

Objective To construct a risk prediction model by integrating the molecular subtypes of pancreatic ductal adenocarcinoma (PDAC) and immune-related genes.Methods With GSE71729 data set (n=145) as the training set,the differentially expressed genes and differential immune-related genes between the squamous and non-squamous subtypes of PDAC were integrated to construct a regulatory network,on the basis of which five immune marker genes regulating the squamous subtype were screened out.An integrated immune score (IIS) model was constructed based on patient survival information and immune marker genes to predict the clinical prognosis of PDAC patients,and its predictive performance was tested with 5 validation sets (n=758).Results PDAC patients were assigned into high risk and low risk groups according to the IIS.In both training and validation sets,the overall survival of patients in the high risk group was shorter than that in the low risk group (both P<0.001).The multivariable Cox regression showed that IIS was an independent prognostic factor for PDAC (HR=2.16,95%CI=1.50-3.10,P<0.001).Conclusion IIS can be used for risk stratification of PDAC patients and may become a potential prognostic marker for PDAC.