ISSN: 0970-938X (Print) | 0976-1683 (Electronic)

Biomedical Research

An International Journal of Medical Sciences

Abstract

Integrated Cox model for survival analysis and biomarker discovery with a feature ranking technique based on z-score transformation in non-small cell lung cancer patients

Purpose: The Purpose of this study was to identify signature biomarker miRNAs that are highly correlated with survival time and other clinical phenotypes. We have done the survival analysis with Cox proportional hazards model and presented a novel technique to identify signature biomarkers.

Materials and Methods: The Data Set from Yu et al. consists of microarray miRNA expression profiles of 112 Non-Small Cell Lung Cell (NSCLC) patients. There are 55 Adenocarcinoma, 50 Squamous cell carcinoma and 7 large cell carcinoma patients. Among them 24 are females and 88 are males, 38 dead and 74 alive. Also it has Stage-I-47, Stage-II-28 and Stage-III-37 patients.

Results: This leads to the efficient calculation of risk score and the patients who are designated as high risk patients. It deals with a novel biomarker selection/ranking technique zsTrans, through which the highly discriminant biomarkers are identified for various histological subtypes such as adenocarcinoma, and stages such as Stage-I, Stage-II and Stage-III of Non-Small Cell Lung Cancer (NSCLC) patients.

Conclusion: The identified Biomarker miRNAs play vital role in the diagnosis and prognosis of NSCLC patients.

Author(s): Anidha M, Premalatha K
Abstract | Full-Text | PDF

Share this  Facebook  Twitter  LinkedIn  Google+