Comparability of Microarray Technology
Author Information
Author(s): Shi Leming, Tong Weida, Fang Hong, Scherf Uwe, Han Jing, Puri Raj K, Frueh Felix W, Goodsaid Federico M, Guo Lei, Su Zhenqiang, Han Tao, Fuscoe James C, Xu Z Alex, Patterson Tucker A, Hong Huixiao, Xie Qian, Perkins Roger G, Chen James J, Casciano Daniel A
Primary Institution: National Center for Toxicological Research, U.S. Food and Drug Administration
Hypothesis
Is the cross-platform comparability of microarray technology reliable?
Conclusion
The study found that low cross-platform concordance is mainly due to low intra-platform consistency and poor data analysis methods, rather than inherent technical differences among platforms.
Supporting Evidence
- The study reanalyzed a dataset and found low intra-platform consistency.
- Different gene selection methods were tested, showing varying levels of cross-platform concordance.
- The results suggest that data analysis methods significantly impact the reliability of microarray results.
Takeaway
This study looked at how well different microarray machines agree with each other. It found that the way the data is analyzed can make a big difference in the results.
Methodology
Reanalysis of Tan's dataset using different gene selection methods and data filtering procedures.
Potential Biases
Potential bias due to the choice of data analysis methods and the quality of the original dataset.
Limitations
The original dataset had low intra-platform consistency and was analyzed with poor methods.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website