Machine learning-based coagulation-related signature for predicting the prognosis and therapy benefits in lung adenocarcinoma

基于机器学习的凝血相关特征预测肺腺癌的预后和治疗获益

阅读:2

Abstract

BACKGROUND: Venous thromboembolism was a one of frequent complications of lung adenocarcinoma (LUAD) and associated with poor clinical outcome. The interaction of coagulation and tumor immune microenvironment could regulate tumor progression and even affect tumor immune response Limited systematic studies regarding the role of coagulation-related genes (CRGs) in LUAD had been performed. METHODS: Consensus clustering analysis was conducted to distinguish CRGs-related clusters with distinct characters. CRGs-related prognostic signature was developed with LASSO algorithm in TCGA cohort and verified using GSE30219 and GSE31210 cohort. RESULTS: A total of two CRGs-related clusters were identified in LUAD. LUAD patients in cluster 1 was associated with favorable clinical outcome, higher tumor microenvironment score, abundant immune cell infiltration, higher expression of HLA-related genes and immune checkpoints, higher immunophenotype score and low IC50 score. Based on 20 CRGs, we also developed and verified a prognostic signature, which had a better performance in prognosis prediction of LUAD patients compared with other eight existing models. Based on risk score and other prognostic factors, we then constructed a survival prediction nomogram that had a good potential for clinical application. CONCLUSION: We identified two molecular subtypes and a prognostic signature for LUAD based on CRGs. This stratification could help the prognosis prediction and chemotherapy, targeted therapy and immunotherapy strategy guidance of LAUD patients.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。