Research groups form an increasingly diverse range of fields, are becoming involved in the task of designing, gathering and analyzing gene expression data produced by microarray experiments. Although large data sets were generated in the last years, little attention has been given to the statistical requirements of the analysis of such data. Most research done with gene expression data has focused on the development of visualization tools, and standard statistical methods such as cluster analysis and principal components have been applied. These techniques have been useful to summarize information, to identify clusters or groups of genes based on similarity or dissimilarity, and to predict biochemical and physiological pathways for some uncharacterized genes. However, important issues such as experimental design, number of replicates and the power of detecting change of expression have not received much attention if any. Comparison of gene expression patterns of tissues or cells under several conditions provides important information to answer several biological questions. Using the simple fold changes in expression based on the ratio of intensities in the red and green channels, as it has been done in the earlier days, is unreliable and inefficient (Pan et al., 2001). Gene expression data is a noisy one and the challenge now is to design and develop methods that allow the detection of the genuine changes.
Proceedings of the World Congress on Genetics Applied to Livestock Production, Volume 2002. Session 16, , 16.12, 2002
|Download Full PDF||BibTEX Citation||Endnote Citation||Search the Proceedings|
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.