Stability of Gene Selection in Microarray Data Analysis
Author Information
Author(s): Qiu Xing, Xiao Yuanhui, Gordon Alexander, Yakovlev Andrei
Primary Institution: University of Rochester
Hypothesis
The variability of gene selection procedures can be assessed using resampling techniques.
Conclusion
Resampling techniques can help identify stable genes and assess the variability of gene selection methods.
Supporting Evidence
- Resampling techniques can reduce the set of initially selected genes to those with high selection frequency.
- Stability properties of multiple testing procedures were described in detail.
- Simulation studies showed the effect of correlation on gene selection performance.
Takeaway
This study looks at how often certain genes are picked in tests to see if they are different from others, helping scientists find the most reliable genes.
Methodology
The study used resampling techniques, including the delete-d-jackknife method, to assess gene selection stability.
Potential Biases
The study does not explicitly mention bias risks.
Limitations
The study's findings may not directly apply to biological data due to the nature of simulations used.
Participant Demographics
The study analyzed data from 335 subjects with childhood leukemia, balanced into two groups of 43 patients each.
Statistical Information
P-Value
3.928 × 10-6
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website