Machine learning reveals diverse cell death patterns in lung adenocarcinoma prognosis and therapy

Machine learning reveals diverse cell death patterns in lung adenocarcinoma prognosis and therapy

2024 | Shun Wang, Ruohuang Wang, Dingtao Hu, Caoxu Zhang, Peng Cao, Jie Huang
This study investigates the role of programmed cell death (PCD) pathways in lung adenocarcinoma (LUAD) prognosis and therapy. The authors identified 13 PCD patterns and integrated bulk RNA, single-cell RNA transcriptomics, and clinicopathological data from TCGA-LUAD and six GEO datasets. They developed a machine learning-based model to identify ten critical differentially expressed genes (CHEK2, KRT18, RRM2, GAPDH, MMP1, CHRNA5, TMPRSS4, ITGB4, CD79A, and CTLA4) associated with PCD in LUAD. A programmed cell death index (PCDI) was constructed using these genes, and it was validated across multiple cohorts. The PCDI was found to correlate with immune features, such as immune cell infiltration and immune checkpoint molecule expression, and was associated with patient outcomes. Patients with a high PCDI score showed resistance to immunotherapy and standard adjuvant chemotherapy but may benefit from other FDA-approved drugs like docetaxel and dasatinib. The study also developed prognostic nomograms based on the PCDI and clinical features, which accurately predicted survival outcomes. The findings suggest that the PCDI has potential as a prognostic biomarker and can guide personalized treatment strategies for LUAD patients.This study investigates the role of programmed cell death (PCD) pathways in lung adenocarcinoma (LUAD) prognosis and therapy. The authors identified 13 PCD patterns and integrated bulk RNA, single-cell RNA transcriptomics, and clinicopathological data from TCGA-LUAD and six GEO datasets. They developed a machine learning-based model to identify ten critical differentially expressed genes (CHEK2, KRT18, RRM2, GAPDH, MMP1, CHRNA5, TMPRSS4, ITGB4, CD79A, and CTLA4) associated with PCD in LUAD. A programmed cell death index (PCDI) was constructed using these genes, and it was validated across multiple cohorts. The PCDI was found to correlate with immune features, such as immune cell infiltration and immune checkpoint molecule expression, and was associated with patient outcomes. Patients with a high PCDI score showed resistance to immunotherapy and standard adjuvant chemotherapy but may benefit from other FDA-approved drugs like docetaxel and dasatinib. The study also developed prognostic nomograms based on the PCDI and clinical features, which accurately predicted survival outcomes. The findings suggest that the PCDI has potential as a prognostic biomarker and can guide personalized treatment strategies for LUAD patients.
Reach us at info@study.space