Machine learning-based immune prognostic model and ceRNA network construction for lung adenocarcinoma
Xiaoqian He1, Ying Su1, Pei Liu1
1College of Information Science and Engineering, Xinjiang University, Urumqi, 830046, China.
Journal of Cancer Research and Clinical Oncology
|March 20, 2023
Summary
This study identifies five immune-related genes (ADM2, CDH17, DKK1, PTX3, AC145343.1) as potential biomarkers for lung adenocarcinoma (LUAD). A novel prognostic model using these genes predicts overall survival, aiding immunotherapy strategies for LUAD patients.
Area of Science:
- Oncology
- Immunology
- Bioinformatics
Background:
- Lung adenocarcinoma (LUAD) has a high mortality rate, necessitating improved treatment strategies.
- Immunotherapy offers a promising approach to enhance patient survival and prognosis in LUAD.
- Identifying novel immune-related markers is crucial for advancing LUAD treatment.
Purpose of the Study:
- To screen reliable immune-related markers for LUAD using bioinformatics and machine learning.
- To construct a prognostic model for predicting overall survival (OS) in LUAD patients.
- To promote the clinical application of immunotherapy in LUAD.
Main Methods:
- Utilized The Cancer Genome Atlas (TCGA) database with 535 LUAD and 59 healthy samples.
- Employed bioinformatics and Support Vector Machine Recursive Feature Elimination to identify Hub genes.
- Constructed a multifactorial Cox regression-based immune prognostic model and nomogram for OS prediction.
Main Results:
- Five immune-related genes (ADM2, CDH17, DKK1, PTX3, AC145343.1) were identified in LUAD.
- ADM2 and AC145343.1 showed a good prognosis (HR<1), while the others indicated a poor prognosis (HR>1).
- Patients in the low-risk group demonstrated significantly better OS rates (P<0.001).
Conclusions:
- Developed an immune prognostic model to predict OS in LUAD patients.
- Highlighted the correlation between five immune genes and immune cell infiltration levels.
- Provided novel biomarkers and insights for LUAD immunotherapy.
