Skip to main content
Fig. 1 | BMC Pulmonary Medicine

Fig. 1

From: A functional genomic model for predicting prognosis in idiopathic pulmonary fibrosis

Fig. 1

Correlation of gene co-expression modules with clinical traits in training cohort (n = 45). Gene co-expression modules were constructed using R package WGCNA (see Methods and Additional file 1 for detail), and denoted by different colors. The parameters for topological overlap matrix generation and unsupervised gene clustering are displayed in Additional file 3: Figure S2A & S2B. The number of genes in each gene module is labeled on left. The module eigengene is the principal component of each gene module computed across all samples. Correlation of module eigengene with each clinical trait was determined by Pearson’s correlation algorithm and displayed in the corresponding box (coefficient on top and p-value in parenthesis on bottom). The color of each box represents the direction of correlation (red) or anti-correlation (green) and the degree of correlations are scaled by the bar on the right. Traits significantly associated with specific modules are highlighted with a purple frame. FVC % predicted = forced vital capacity percent predicted; DLCO % predicted = diffusion capacity of carbon monoxide percent predicted; CPI = composite physiologic index

Back to article page