We evaluated the performance of various centrality measures by comparing the area under the ROC curves and the minimum distance of the ROC curves from the ideal optimum classification point where TPR (True Positive Rate) is 1 and FPR (False Positive Rate) is 0. We further investigated that the functions of yeast proteins also have some relations with their centrality measures. Different types of centrality values imply different types of importance of a node in a network.
By deeply examining different centrality values of yeast proteins we found that they are not highly correlated, which leaded us to hypothesize that centralities might have some relations with gene/protein functionalities. Indeed, we found that many of the clusters generated based on the pattern of centrality values are rich with similar function proteins. The statistical significance of the protein clusters was assessed by hyper-geometric p-values. Using the statistically significant clusters, we established links between pattern of centrality measures and protein functions.