Fig. 2.

Evaluation scheme. To evaluate CCI methods, 90 simulated datasets and six real empirical datasets were prepared. Four ligand–receptor (L–R) scoring methods and six binarization methods were then evaluated. For the evaluation of these methods, area under the receiver operating characteristic curve (AUCROC), area under the precision–recall curve (AUCPRC), memory usage, and computational time were determined. For the evaluation of binarization methods, F-measure, Matthews correlation coefficient (MCC), positive ratio (PR), false positive ratio (FPR), and false negative ratio (FNR) were determined