Related Experiment Videos
Testing association of a pathway with survival using gene expression data.
Jelle J Goeman1, Jan Oosting, Anne-Marie Cleton-Jansen
1Department of Medical Statistics, Leiden University Medical Center, The Netherlands. j.j.goeman@lumc.nl
Bioinformatics (Oxford, England)
|January 20, 2005
Summary
A new score test extends the Global Test methodology for analyzing gene expression and patient survival data. This method directly assesses the impact of gene groups on survival, aiding in the discovery of relevant biological pathways.
Area of Science:
- Biostatistics
- Bioinformatics
- Genomics
Background:
- Growing interest in survival analysis for microarray studies necessitates extending existing statistical methodologies.
- The Global Test, a powerful tool for analyzing high-dimensional data, requires adaptation for survival endpoints.
Purpose of the Study:
- To develop and present a novel score test for assessing the association between gene expression profiles and survival time.
- To enable direct hypothesis testing on the influence of gene groups (e.g., pathways, genomic regions) on survival, bypassing single-gene analysis.
Main Methods:
- The proposed test is based on the Cox proportional hazards model.
- Calculations utilize martingale residuals, allowing for direct assessment of gene group effects on survival.
- The methodology incorporates adjustments for covariates and includes a diagnostic graph for result interpretation.
Main Results:
- A score test for the association between gene expression profiles and survival time is presented.
- The test effectively evaluates the influence of predefined gene groups on patient survival.
- Application to a tumor dataset identified specific gene ontology pathways associated with patient survival.
Conclusions:
- The developed score test provides a direct and efficient method for analyzing the relationship between gene expression patterns and survival.
- This approach facilitates the identification of biologically relevant pathways influencing patient outcomes.
- The methodology is implemented in the R-package 'globaltest', enhancing its accessibility for researchers.