Related Experiment Video
Updated: Aug 6, 2026
07:41
Performing Data Mining And Integrative Analysis Of Biomarker in Breast Cancer Using Multiple Publicly Accessible Databases
Published on: May 17, 2019
8.9K
Leveraging Bioinformatics and Machine Learning for Identifying Prognostic Biomarkers and Predicting Clinical Outcomes
Kaida Cai1,2,3, Wenzhi Fu2, Hanwen Liu2
1Department of Epidemiology and Biostatistics, School of Public Health, Southeast University, Nanjing 210009, China.
Genes
|January 8, 2025
Summary
Researchers identified DKK1 and TNS4 as key genetic predictors for lung adenocarcinoma (LUAD) survival. High expression correlates with poorer outcomes, aiding personalized treatment strategies.
Area of Science:
- Oncology
- Genomics
- Bioinformatics
Background:
- Lung adenocarcinoma (LUAD) presents significant challenges due to poor prognosis and limited advanced-stage treatment options.
- Identifying genetic biomarkers is crucial for improving outcome prediction and guiding personalized therapies in LUAD.
Purpose of the Study:
- To identify key genetic features from ultra-high dimensional RNA-sequencing data in LUAD patients.
- To evaluate and compare the predictive performance of different survival analysis methods for LUAD.
- To explore the biological significance of identified genes using protein-protein interaction networks.
Main Methods:
- Utilized a multi-step approach combining sure independence screening, penalized regression, and information gain for feature selection.
- Evaluated Cox model, survival tree, and random survival forests (RSFs) for survival analysis.
- Constructed a protein-protein interaction network to analyze gene relationships.
Main Results:
- DKK1 and TNS4 were consistently identified as significant predictors across all feature selection methods.
- High expression of DKK1 and TNS4 correlated with poorer survival outcomes (Kaplan-Meier analysis).
- RSF demonstrated superior predictive performance (higher AUC and C-index) compared to Cox and survival tree methods. VEGFC and LAMA3 emerged as key nodes in the protein-protein interaction network.
Conclusions:
- DKK1 and TNS4 show potential as prognostic biomarkers for LUAD.
- RSF is a highly effective method for survival prediction in LUAD.
- The identified genes and network provide insights into LUAD's genetic mechanisms, supporting the development of prognostic tools and personalized treatments.
