Fig. 3

Regulatory divergence patterns of duplicate genes in co-expression network partitioning. a Classification of different types of co-expression patterns for duplicate genes. The proportion of shared genes (x axis and y axis) of a specific gene module is defined as the number of shared correlated genes between two duplicates divided by the total number of correlated genes for one of the duplicate genes. Each dot represents a pair of duplicate genes. A total of six co-expression regulatory patterns were classified and an example of the type of networks is visualized in the pink boxes along with the number of genes in each pattern. The other duplicate pairs were designated as “unclassified”. b The enrichment or depletion of genes in each co-expression pattern was assessed for each of the categories of gene duplications. The log₂ ratio of observed to expected proportion of genes in each class is shown and “**” are used to indicate significant enrichment (P < 0.01). The expectation was calculated based on the proportion of duplicate pairs shown in Fig. 2b. c The co-expressed divergence patterns was related to the age (indicated by dS) of the duplications. The synonymous mutation rate (dS) divided by the standard deviation of dS among WGD pairs is shown for each class of co-expression relationships