Construction method of prediction model for brain glioma prognosis

The invention discloses a construction method of a prediction model for brain glioma prognosis. The construction method comprises the following steps: acquiring data of a gene expression database; screening out differential expression genes related to pyroptosis according to the data; screening out key genes based on differential expression genes; constructing a prognosis risk assessment model according to the key genes; and inputting the prognosis risk assessment model into a verification database, and verifying a prediction effect. According to the application, a group of pyroptosis related genes are used for overcoming dispute diagnostic criteria, variability among observers and non-representative samples, and inaccurate survival evaluation of glioma patients of the same level is avoided. 本申请公开了一种用于脑胶质瘤预后的预测模型的构建方法，步骤包括：获取基因表达数据库的数据；根据数据筛选出与细胞焦亡相关的差异表达基因；基于差异表达基因筛选出关键基因；根据关键基因构建预后风险评估模型；将预后风险评估模型输入验证数据库中，验证预测效果。本申请通过一组细胞焦亡相关基因用来克服有争议的诊断标准、观察者间的变异性和不具代表性的样本，避免了相同级别的胶质瘤患者的不准确的生存评估。