Regularized binormal ROC method in disease classification using microarray data
Author Information
Author(s): Ma Shuangge, Song Xiao, Huang Jian
Primary Institution: University of Washington
Hypothesis
Can the binormal AUC be effectively used for disease classification and biomarker selection in high-dimensional genomic data?
Conclusion
The proposed approach is computationally more affordable without losing the optimality possessed by the standard ROC method.
Supporting Evidence
- The proposed method showed satisfactory classification performance with small classification errors and large AUCs.
- Simulation studies demonstrated the method's robustness under various scenarios.
- Application to cancer datasets indicated biologically sound implications of identified genes.
Takeaway
This study shows a new way to find important genes for disease diagnosis using a method that is faster and still accurate.
Methodology
The study used simulation studies and applied the method to two cancer datasets, employing V-fold cross validation for tuning parameter selection.
Potential Biases
Potential biases may arise from the selection of genes based on adjusted t-statistics.
Limitations
The method's theoretical construction is based on the normal distribution assumption, which may not hold in all cases.
Participant Demographics
The study involved cancer datasets with gene expression data from tumor and normal tissues.
Statistical Information
P-Value
p<0.0001
Statistical Significance
p<0.0001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website