Figure 3
From: An integer optimization algorithm for robust identification of non-linear gene regulatory networks
Results from 5-gene network identified under data uncertainty with 5% noise. (a) Number of bootstrap occurrences for each connection (1000 bootstrap samples total). (b) Identified network structure. Numbers above each connection represent percent occurrence, with the thick lines representing the number of connections appearing in more than 90% of the bootstrapped samples and the thin lines representing the connections appearing in more than 45% of bootstrapped samples. (c) frequency of specific connection values shown as heat map.