Skip to main content
. 2018 Sep 18;8:13995. doi: 10.1038/s41598-018-32275-8

Figure 3.

Figure 3

Hierarchical clustering and network visualisation. The cytokine and PAI-1 values were transformed to common logarithm values to normalise the distribution of the data. (A) Hierarchical clustering of Pearson’s correlations between cytokines and PAI-1. The outlined boxes show the common cytokine network in the acute phase of sepsis (days 1–4). (B) This network visualises the significant correlations of (A). The cytokines and PAI-1 with log2 fold change (i.e. average cytokine levels in sepsis patients/average cytokine levels in controls) > 1.5 were included. The size of each node was determined based on the log2 fold change. Red, yellow, and grey colours indicate a significant increase in the cytokines and PAI-1 levels between septic patients and controls, between critically ill patients and controls, and no difference between septic or critically ill patients and controls, respectively. Increased cytokines compared to those of the controls and the connections are shown to represent networks with major impact. IFN-α: interferon-α; IFN-γ: interferon-γ; IL-1β: interleukin-1 beta; IL-6: interleukin-6; IL-8: interleukin-8; IL-12/IL-23p40: interleukin-12/23p40; IL-17A: interleukin-17A; TNF-α: tumour necrosis factor-α; MCP-1: monocyte chemotactic protein-1; IL-4: interleukin-4; IL-10: interleukin-10; PAI-1: plasminogen activator inhibitor-1.