列线图
接收机工作特性
医学
肿瘤科
单变量
比例危险模型
基因签名
逻辑回归
内科学
多元统计
基因
生物
统计
基因表达
遗传学
数学
标识
DOI:10.1016/j.archoralbio.2022.105375
摘要
The present study aimed to construct prognostic and diagnostic gene signatures and nomograms in oral carcinoma.Differentially expressed genes between oral carcinoma and normal samples were analyzed based on data from The Cancer Genome Atlas and Gene Expression Omnibus. Cox univariate and multivariate analyses and lasso analysis were conducted to construct the prognostic gene signature. Moreover, logit regression analysis was performed to establish the diagnostic model. Receiver operating characteristic curve, nomogram, calibration curve were utilized for validating the performance of the signature and the diagnostic model.A new 4-gene (ZNF114, AREG, GRB14, and DDIT4) prognostic signature was successfully constructed. Samples with higher risk score exhibited a worse overall survival than that with lower risk score in both The Cancer Genome Atlas cohort and GSE41613. Moreover, the receiver operating characteristic curve revealed a good performance of this signature in the prediction of survival. Furthermore, the signature was an independent prognosticator for oral carcinoma patients. Also, the nomogram and calibration curves showed a good clinical prediction effect. What's more, a diagnostic model including ZNF114, GRB14, and DDIT4 was constructed and presented a good performance in distinguishing oral carcinoma and normal samples.The present study developed a new 4-gene prognostic signature and nomogram, and constructed a 3-gene diagnostic model and nomogram to help in the diagnosis and prognosis of oral carcinoma.
科研通智能强力驱动
Strongly Powered by AbleSci AI