Skip to main content
. 2023 Apr 2;39(4):btad158. doi: 10.1093/bioinformatics/btad158

Figure 1.

Figure 1.

The workflow of our method. The initial sketch is created by combining literature-based knowledge about the model and possible initial hypotheses about the model. The sketch may be further complemented with properties obtained automatically from binarized data, in which case, we also call it the data-informed sketch. Once we run the BN inference procedure, we obtain a set of BN candidates, which we may enumerate and further inspect. This may lead us to the formulation of new hypotheses or suggest new experiments to perform, both of which may further be used to refine the sketch and run the procedure again