Advertisement

< Back to Article

Modulated Modularity Clustering as an Exploratory Tool for Functional Genomic Inference

Figure 2

Correlation structure for simulated data.

Twelve variables are correlated as shown in the figure. The heat map used to illustrate the pairwise correlations ranges from dark red (perfect correlation, ) through green (no correlation, ) to dark blue (perfect anticorrelation, ). On the diagonal are four correlated clusters of varying strength: {1,2} with an of 0.9, {3,4,5,6} with an of 0.7, {7,8,9} with an of 0.6, and {10,11,12} with an of 0.8. There are nonzero correlations between two pairs of clusters; members of {1,2} and {3,4,5,6} are positively correlated (), while members of {7,8,9} and {10,11,12} are negatively correlated ().

Figure 2

doi: https://doi.org/10.1371/journal.pgen.1000479.g002