Skip to main content
Figure 5 | BMC Bioinformatics

Figure 5

From: Usefulness and limitations of dK random graph models to predict interactions and functional homogeneity in biological networks under a pseudo-likelihood parameter estimation approach

Figure 5

The performance of dK distribution models in predicting functionally homogeneous modules based on GDS1013 coexpression data (PCC cut-off threshold 0.89). The gene group size was 10, the p-value cut-off threshold was 10-5, and p = 0.9. a) Accuracy; b) ROC curve; c) Precision-Recall. Note that the curve for the 0K model is truncated near (0.80,0.89); this is due to an abundance of modules having the maximum density, and no way to distinguish between them in the 0K measure. Therefore, the minimum recall computable is that shown in the graph.

Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us