Figure 7

ROC curves of TF-gene regulatory network reconstruction for the synthetic data with MFA and BFA. This figure shows various receiver operating characteristic (ROC) curves, where the numbers of predicted true positive interactions (vertical axis) are plotted against the numbers of false positive interactions (horizontal axis). Larger areas under the curve (AUC) indicate a better reconstruction accuracy. (a), (b) show the ROC curves obtained from TF binding data alone, without including gene expression profiles. (a) corresponds to the noisy TF binding data shown in Figure 4(c). (b) corresponds to the less noisy TF binding data, shown in Figure 4(d). (c), (d) each composed of 9 graphs, show the predictions obtained with MFA-VBEM from both noisy TF binding and gene expression profiles. (e), (f) also composed of 9 graphs each, show the results obtained with BFA-Gibbs on the same data. The arrangement of the graphs is the same as in Figure 6. The results suggest that MFA-VBEM systematically outperforms BFA-Gibbs. They also suggest that for noisy TF binding data (c), (e), the inclusion of gene expression profiles and the application of MFA-VBEM leads to an improvement in the TF-gene regulatory network reconstruction.