Comparative performance of seriation and PoissonC on a simulated SAGE data set with 10 expression patterns.

Seriation and PoissonC were applied to the analysis of a simulated SAGE data set containing 10 expression patterns each including 50 tags, and 50 noise tags. TP (True Positives) are tags that were correctly classified as belonging to the right expression pattern or noise. FP (False Positives) are tags that were assigned to the wrong pattern or noise tags that were assigned to an expression pattern.
*The false positive rate is significantly higher for the PoissonC algorithm than it is for seriation (p<0.05).