Skip to main content


Figure 5 | BMC Evolutionary Biology

Figure 5

From: Efficient context-dependent model building based on clustering posterior distributions for non-coding sequences

Figure 5

Graphical representation of the graph-based context reduction approach. The graphical representation of the graph-based reduction approach illustrates the coloring scheme to build the different clusters (the complete graph figure was obtained using GrInvIn; see [34]). The edges are labeled with the step of the graph-based algorithm during which they were colored. Given the large number of vertices in the graph, there are many possibilities of merging contexts into clusters. The coloring of nodes and vertices in this figure reveals an optimum of 8 clusters for the Ancestral Repeats dataset: 5 separate dark grey clusters (each containing a single context: AXA, AXG, CXT, TXA and TXT) and 3 colored clusters. The green cluster contains 4 contexts: AXC, GXA, GXG and TXG, the yellow cluster contains 2 contexts: AXT and CXG, and the red cluster contains 5 contexts: CXA, CXC, GXC, GXT, and TXC.

Back to article page