Figure 4
From: Nearest Neighbor Networks: clustering expression data based on gene neighborhoods
Functional evaluation of clustering algorithms. Function-specific evaluation results for each clustering method on a per data set and GO term basis. Each cell represents an AUC score calculated analytically using the Wilcoxon Rank Sum formula; below baseline performance appears in blue, and yellow indicates higher performance. Data set and term combinations for which ten or fewer pairs were able to be evaluated are excluded and appear as gray missing values; functions for which less than 10% of methods were available due to gene exclusion by NNN, QTC, or SAMBA were removed. Visualization provided by TIGR MeV [41].