Figure 3From: Improving the sensitivity of sample clustering by leveraging gene co-expression networks in variable selectionModule structure in the gene co-expression network from the simulated dataset of d = 500. (a). Two modules were identified in the co-expression network. The rest on the right are zoomed-in view of the modules highlighting the genes included in two modules respectively. (b) and (c), hard thresholding transformation, with top 1% and 5% correlations included in the network, respectively. (d) and (e), soft thresholding transformation, with the power functions parameter β=3 and β=7, respectively.Back to article page