Skip to main content
Figure 4 | Genome Biology

Figure 4

From: Functional classification of proteins for the prediction of cellular function from a protein-protein interaction network

Figure 4

Robustness of PRODISTIN towards false interactions. The prediction rate (number of correct predictions divided by number of predictions) was measured for PRODISTIN (yellow curve) and for the majority rule algorithm (green curve) on networks on which a certain percentage of interactions were randomly 'rewired' (from 10 to 50%) (see text). The number of proteins for which a prediction is possible is also reported as a histogram (dark red, PRODISTIN; blue, majority rule). The values correspond to an average of 50 experiments for each percentage of false interactions introduced into the dataset.

Back to article page