GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists

BMC Bioinformatics

Table 1 A comparison of web-based GO enrichment tools.

Tool	P-value and statistical method	Flexible threshold	Graphical visualization	Multiple organisms	Running time
GOrilla	Exact mHG p-value computation (no need for simulations)	+	+	+	7 Sec
Fatiscan [13]	Fischer Exact (FDR corrected for number of thresholds)	+ (predetermined steps of 30)	-	+	30 Min
GO-stat [14]	Wilcoxon Rank-Sum/Kolmogorov Smirnov	+	-	+	2 Min
GOEAST [9]	Hypergeometric	-	+	+	20 Min
SGD [11]	Hypergeometric	-	+	- (only yeast)	2 Min
DAVID [7]	Modified Fischer Exact	-	-	+	2 Min
GOTM [10]	Hypergeometric	-	+	+	2 Min
GoMiner [3]	Fisher Exact	-	- (only in the downloadable version)	+	7 Min

Different GO enrichment tools employ a wide range of statistics and yield different performances. The main features of five different web-based tools are compared to GOrilla. To enable a fair comparison all tools were used using default parameters via their web interfaces and applied on the van't Veer dataset. One exception is the SGD tool that only runs on yeast data and was therefore tested on a set of 543 yeast genes and the default background for running time characterization. The GOrilla running time for this yeast dataset was also 7 seconds. The running time was measured for the entire analysis, including uploading files and getting the results.

ISSN: 1471-2105