FSim: A Novel Functional Similarity Search Algorithm and Tool for Discovering Functionally Related Gene Products

Figure 4

Performance comparison of different algorithms using all pathway data sets. The box-whisker plots show the AUC value distributions of the FSim method using separate BP (FSim_BP) ontology, MF (FSim_MF) ontology, and CC (FSim_CC) ontology and using all three ontologies (FSim), as well as the other five evaluated methods. The band in each box is the median AUC values of each method, which can be used to compare the overall performance of all the methods.