Figure 2From: Improving accuracy of protein-protein interaction prediction by considering the converse problem for sequence representationThe ROC curves of four available predicting methods on Escherichia coli and Saccharomyces cerevisiae datasets. A, ROC curves on Escherichia coli data with negative samples constructed by sub-cellular information; B, ROC curves on Escherichia coli data with negative samples sampled randomly from the complementary network; C, ROC curves on Saccharomyces cerevisiae data with negative samples constructed by subcellular information; D, ROC curves on Saccharomyces cerevisiae data with negative samples sampled randomly from the complementary network.Back to article page