Figure 4

Distinct domain architectures compared to genome graph properties. a) Mean Degree b) Mean Clustering Coefficient c) Characeristic Path Length d) Percentage of pairs with paths. Each genome is represented by one point (labelled with its abbreviation used in the SUPERFAMILY database). Archaea are coloured red, bacteria yellow, eukaryotes blue and groups of genomes (e.g. all archaea) are shown in grey. There is a strong correlation between the mean degree and number of distinct architectures (correlation coefficient 0.99). Clustering coefficient (0.85), characteristic path length (0.73) and percentage of connected pairs (0.88) are also correlated, though more weakly.