Tumor classification ranking from microarray data
2008
Ranking Tumor Classifications from Microarray Data
Sample size: 327
publication
Evidence: moderate
Author Information
Author(s): Rattikorn Hewett, Phongphun Kijsanayothin
Primary Institution: Texas Tech University
Hypothesis
Can tumor classification be made more informative using a ranking method that maintains good classification accuracy?
Conclusion
The MDR algorithm provides effective and informative tumor classifications from cancer gene expression data.
Supporting Evidence
- MDR achieved an average AUC of 91.01% across 11 cancer types.
- MDR outperformed other algorithms in classifying ALL-AML leukemia.
- MDR maintained high accuracy while reducing the number of predictor genes significantly.
Takeaway
This study shows a new way to classify tumors using gene data, making it easier to understand which types of cancer are present.
Methodology
Microarray data for 11 types of cancer were analyzed using the Multi-Dimensional Ranker (MDR) algorithm.
Limitations
The study is preliminary and validated on a limited number of data sets.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website