Fig. 5
From: A novel method to predict essential proteins based on tensor and HITS algorithm
PR curves of HEPT and other existing centrality methods. The scoring of proteins in top K (cutoff value) are chosen as candidate essential proteins and the residual proteins in PPI network are seen as candidate nonessential proteins by each method (HEPT, DC, IC, SC, BC, CC, NC, PeC, CoEWC, and POEM). With choosing different values of K, the results of precision and recall are calculated for each method respectively. The results of precision and recall are plotted in PR curves with different cutoff values. a The PR curves of HEPT, NC, DC, IC, SC, BC, and CC. b The PR curves of HEPT and other three methods: CoEWC, PeC, and POEM