TABLE 6.
Results from testing global network metrics and centrality measures of the networks in Figure 5B.
| CMC group | Control group | abs.diff. | |
| Global network measures: | |||
| Average path length [1] | 1.996 | 1.907 | 0.089 |
| Clustering coefficient [2] | 0.294 | 0.329 | 0.035 |
| Modularity [3] | 0.486 | 0.531 | 0.045 |
| Vertex connectivity [4] | 1.000 | 1.000 | 0.000 |
| Edge connectivity [5] | 1.000 | 1.000 | 0.000 |
| Density [6] | 0.131 | 0.118 | 0.013 |
| Degree [7]: | |||
| s__Roseburia_sp._CAG.18 (X138) | 8 | 3 | 5 |
| s__Clostridium_sp._CAG.7 (X44) | 5 | 1 | 4 |
| s__Prevotella_sp._CAG.386 (X530) | 4 | 6 | 2 |
| s__Bacteroides_stercoris (X3) | 1 | 3 | 2 |
| s__Bacteroides_dorei (X14) | 3 | 5 | 2 |
| Betweenness centrality [8]: | |||
| s__Prevotella_sp._CAG.386 (X530) | 4 | 113 | 109 |
| s__unclassified (X1) | 9 | 108 | 99 |
| s__Bacteroides_unclassified (X2) | 57 | 153 | 96 |
| s__Roseburia_sp._CAG.18 (X138) | 98 | 3 | 95 |
| s__Faecalibacterium_prausnitzii (X9) | 34 | 193 | 69 |
| Closeness centrality [9]: | |||
| s__Bacteroides_stercoris (X3) | 2.885 | 19.713 | 16.827 |
| s__Bacteroides_stercoris_CAG.120 (X4) | 2.885 | 16.422 | 13.537 |
| s__Prevotella_sp._CAG.386 (X530) | 19.32 | 25.916 | 6.597 |
| s__Clostridium_sp._CAG.7 (X44) | 21.449 | 15.646 | 5.804 |
| s__unclassified (X1) | 20.144 | 25.228 | 5.084 |
| Eigenvector centrality [10]: | |||
| s__Clostridium_sp._CAG.7 (X44) | 0.158 | 0.010 | 0.147 |
| s__Ruminococcus_gnavus (X117) | 0.185 | 0.052 | 0.133 |
| s__Bacteroides_dorei (X14) | 0.171 | 0.293 | 0.122 |
| s__Bacteroides_massiliensis (X5) | 0.108 | 0.227 | 0.119 |
| s__Bacteroides_stercoris (X3) | 0.034 | 0.150 | 0.116 |
Shown are, respectively, the computed measure for CMC group and control group, the absolute difference between groups was computed; for degree, betweenness centrality, closeness centrality, and eigenvector centrality analysis: the five taxa with the highest absolute group difference are shown. Local and global network properties implemented in NetCoMi: [1] Arithmetic mean of all shortest paths between vertices in a network. [2] Proportion of triangles with respect to the total number of connected triples2, Expresses how likely the nodes are to form clusters. [3] Expresses how well the network is divided into communities (many edges within the identified clusters and only a few between them). [4][5] Minimum number of edges or vertices (nodes) that need to be removed to disconnect the network, respectively. Not meaningful for a fully connected network. [6] Ratio of the actual number of edges in the network and the possible number of edges. Not meaningful for a fully connected network. [7] Number of adjacent nodes. [8] Fraction of times a node lies on the shortest path between all other nodes. A central node has the ability to connect sub-networks. [9] Reciprocal of the sum of shortest paths between this node and all other nodes. The node with the highest closeness centrality has the minimum shortest path to all other nodes. [10] Calculated via eigenvalue decomposition: Ac = λc, where λ denotes the eigenvalues and c denotes the eigenvectors of the adjacency matrix A. Eigenvector centrality is then defined as the i-th entry of the eigenvector belonging to the largest eigenvalue A node is central if it is connected to other nodes having themselves a central position in the network.