Figure 5.

Preterm infant blood proteomic analysis using weight gene correlation network analysis (WGCNA) and identification of LCN2

(A–G) WGCNA identified 15 modules at day 7 (A, top panel) among which 5 modules were significantly correlated with CRP (B, top panel). At day 14, 16 modules were identified (A, bottom panel), among them 6 modules were significantly correlated with CRP (B, bottom panel). The numbers indicate the r score and in brackets the p value. (C) A total of 231 proteins identified at day 7 and 144 proteins identified at day 14 were associated with CRP. The intersection of the significant modules identified 100 proteins as common for both time points. (D) Protein-protein interaction (PPI) network constructed from the 100 overlapping proteins between the highly correlated proteins with CRP at day 7 and 14. (E) Enrichment map of significant gene ontology (GO) terms. Nodes represent gene sets. Highly similar protein sets are connected by edges, grouped in sub-clusters, and annotated automatically. (F) The top 10 high-degree proteins screened from the PPI network by using Maximal Clique Centrality (MCC) algorithm. The network was created using Cytohubba plug-in in Cytoscape. (G) LCN2 was positively correlated with CRP at both day 7 and day 14. See also Figure S3 and Tables S3 and S4.