Skip to main content


Fig. 4 | BMC Bioinformatics

Fig. 4

From: Efficient randomization of biological networks while preserving functional characterization of individual nodes

Fig. 4

CellNOpt study case. a Analysis of the Jaccard index trend across switching-steps (SS) while rewiring the two bipartite network induced by the positive (respectively negative) edges of the reference DSN (liver prior knowledge network (liver-PKN)) and estimation of the lower bounds for the number of switching-steps; b visual inspection of the switching-algorithm Markov chain convergence to verify the suitability of the estimated bounds (see Fig. 2 legend for further details); c Comparison of the CellNOpt scores and the rewired scores; d Empirical p-values of the CellNOpt scores across the entire family of models. e The liver-PKN used by CellNOpt as initial reference network; f The model outputted by CellNOpt when using the liver-PKN as initial reference network with superimposed the frequency of inclusion of each node in a set of 1,000 models outputted by CellNopt using F-rewired versions of the liver-PKN as reference networks

Back to article page