Finding significance in array data

Much effort has been applied to developing software for analyzing the results of DNA microarray experiments. The majority of this work has been done by biologists who were focused on identifying regulatory pathways by finding co-regulated genes using clustering algorithms, self-organizing maps and other techniques. But cluster analysis does not provide information about the statistical significance of gene expression levels. As microarrays become more widely used for medical applications, knowing the significance of the extent to which a specific gene or group of genes is up or down regulated is becoming increasingly important.

Robert Tibshirani, professor of statistics at Stanford University, and co-workers have now published in the Proceedings of the National Academy of Sciences a statistical analysis method called Significance Analysis of Microarrays (SAM) to identify statistically significant changes in gene expression. ...