Additional file 2: Figure S1. of Properties of Boolean dynamics by node classification using feedback loops in a network

Comparison between groups of NFD and non-NFD genes in random networks with respect to the proportions of essential genes, disease genes, and drug targets. In each subfigure, a set of 100 random networks were generated by rewiring the interactions of the signaling networks such that the in-degree and the out-degree of all nodes are conserved. (A) Result of random networks shuffled from KEGG network. The average numbers of NFD and non-NFD genes were 509 and 1150, respectively. The average proportions of essential genes in the NFD and the non-NFD groups were 0.2849 and 0.2852, respectively. The proportions of disease genes in the NFD and the non-NFD groups were 0.2417 and 0.2435, respectively. The proportions of drug-targets in the NFD and the non-NFD groups were 0.2141 and 0.2122, respectively. (B) Result of random networks shuffled from WANG network. The average numbers of NFD and non-NFD genes were 1544 and 4761, respectively. The proportions of essential genes in the NFD and the non-NFD groups were 0.2390 and 0.2413, respectively. The proportions of disease genes in the NFD and the non-NFD groups were 0.2455 and 0.2472, respectively. The proportions of drug-targets in the NFD and the non-NFD groups were 0.1768 and 0.1769, respectively. Figure S2. Changes in the proportion of functionally important genes over the threshold value of the perturbation-sustainable probability in random networks. In each subfigure, a set of 100 random networks were generated by rewiring the interactions of the signaling network such that the in-degree and the out-degree of all nodes are conserved. Given a threshold value β, the y-axis values indicate the average proportions of essential genes, disease genes, and drug targets over the set of candidate genes whose perturbation-sustainable probability is larger than or equal to β in random networks. (A) Results in random networks shuffled from KEGG network. For a reliable comparison, the maximal β was set to 0.0266 which generates 131 candidate genes on average. (B) Results in random networks shuffled from WANG network. The maximal β was set to 0.0908, which results in 121 candidate genes on average. (ZIP 362 kb)针对随机网络中NFD基因（NFD genes）与非NFD基因（non-NFD genes）两组，就必需基因、疾病基因及药物靶点的占比展开对比分析。在每张子图中，研究人员通过重连信号网络（signaling networks）的相互作用关系，生成100组随机网络，且保证所有节点的入度与出度均保持守恒。(A) 基于KEGG网络重连得到的随机网络结果：NFD基因与非NFD基因的平均数量分别为509与1150；两组中必需基因的平均占比分别为0.2849与0.2852；疾病基因的占比分别为0.2417与0.2435；药物靶点的占比分别为0.2141与0.2122。(B) 基于WANG网络重连得到的随机网络结果：NFD基因与非NFD基因的平均数量分别为1544与4761；两组中必需基因的占比分别为0.2390与0.2413；疾病基因的占比分别为0.2455与0.2472；药物靶点的占比分别为0.1768与0.1769。 补充图S2（Figure S2）：随机网络内功能重要基因的占比随扰动可持续概率阈值的变化情况。在每张子图中，研究人员通过重连信号网络的相互作用关系生成100组随机网络，且保证所有节点的入度与出度均保持守恒。给定阈值β，纵轴数值代表：在随机网络中，扰动可持续概率大于或等于β的候选基因集合内，必需基因、疾病基因及药物靶点的平均占比。(A) 基于KEGG网络重连得到的随机网络结果：为保障对比可靠性，将最大阈值β设为0.0266，此时平均可获得131个候选基因。(B) 基于WANG网络重连得到的随机网络结果：最大阈值β设为0.0908，此时平均可获得121个候选基因。（ZIP压缩包，大小362 kb）