Fig. 2From: Exploring soybean metabolic pathways based on probabilistic graphical model and knowledge-based methodsThe gene size distribution for 16 clusters. The genes were organized into 16 groups using the Expectation-Maximization clustering method. Genes in the same cluster share a highly similar gene expression pattern, while genes from different clusters have low similarity to each other. These 16 gene clusters were used for gene sampling to enlarge the initial, mapped pathways. X-axis denotes the index of clusters; Y-axis denotes the number of genes in each clusterBack to article page