Figure 2
From: Inferring microRNA and transcription factor regulatory networks in heterogeneous data
An example of Bayesian network structure learning. (a) The initial structure corresponding to regulator R1 and target gene G1. This initial structure is created based on target information of the regulator. (b) All possible candidate structures generated from the initial structure. The interaction between R1 and G1 in each candidate structure is evaluated using a scoring function. (c) The presence or absence of an edge between R1 and G1 based on the average score of each case. The candidate structure with higher average score will be chosen.