Skip to main content
Figure 1 | BMC Bioinformatics

Figure 1

From: Learning transcriptional regulatory networks from high throughput gene expression data using continuous three-way mutual information

Figure 1

A schematic view of the network inference procedure for MI3 and control methods. We learn gene regulatory networks in two steps: (1) learn local regulatory network for each of the interesting nodes through an exhaustive search; (2) assemble local networks up into a unified network if needed. In the step (2), we may need to reconcile the conflicting local structures (labeled by *) if there are any, mainly the two way edges and cycles. More details of the procedure are described in Methods part. In this work, the key difference between different methods is the score metric being used rather than the network inference procedure. For a fair comparison between scoring metrics, we simple assemble the local networks up without the reconciliation of conflicts in step (2).

Back to article page