Identification of Subtype-Specific Prognostic Genes for Early-Stage Lung Adenocarcinoma and Squamous Cell Carcinoma Patients Using an Embedded Feature Selection Algorithm

• Published in: PloS one Vol. 10; no. 7; p. e0134630
• Main Author: Tian, Suyan
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 30.07.2015; Public Library of Science (PLoS)
• Subjects: Adenocarcinoma; Adenocarcinoma - diagnosis; Adenocarcinoma - genetics; Algorithms; Analysis; Bioinformatics; Biomarkers; Cancer genetics; Carcinoma, Squamous Cell - diagnosis; Carcinoma, Squamous Cell - genetics; Classification; Computer simulation; Data analysis; Early Detection of Cancer - methods; Gene expression; Genes; Genomes; Hepatitis; Histology; Humans; Identification; Lung cancer; Lung Neoplasms - diagnosis; Lung Neoplasms - genetics; Medical prognosis; Oligonucleotide Array Sequence Analysis; Prognosis; Proportional Hazards Models; Regression analysis; Regression models; Regularization; Regularization methods; RNA; Sensitivity and Specificity; Squamous cell carcinoma; Studies; Survival analysis
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0134630

The existence of fundamental differences between lung adenocarcinoma (AC) and squamous cell carcinoma (SCC) in their underlying mechanisms motivated us to postulate that specific genes might exist relevant to prognosis of each histology subtype. To test on this research hypothesis, we previously proposed a simple Cox-regression model based feature selection algorithm and identified successfully some subtype-specific prognostic genes when applying this method to real-world data. In this article, we continue our effort on identification of subtype-specific prognostic genes for AC and SCC, and propose a novel embedded feature selection method by extending Threshold Gradient Descent Regularization (TGDR) algorithm and minimizing on a corresponding negative partial likelihood function. Using real-world datasets and simulated ones, we show these two proposed methods have comparable performance whereas the new proposal is superior in terms of model parsimony. Our analysis provides some evidence on the existence of such subtype-specific prognostic genes, more investigation is warranted.