Fig.5. Community detection from transcriptomic/PPI integrated data in SARS-CoV-2-infected lung epithelial cells

Fig.5 - We attempted integration of transcriptome data to protein-protein interaction networks with prize-collecting Steiner forest (PCSF) algorithm. PCSF method allows inclusion of non-responding genes (‘Steiner nodes’) to the subnetworks, therefore non-responding genes that connects the differentially expressed genes can also be identified.
PCSF algorithm extracted an active PPI subnetwork of 131 nodes (32 Steiner and 89 terminal nodes) and 143 edges in NHBE cells in response to virus infection. Louvain community detection algorithm identified 10 communities within the active PPI subnetwork.
Steiner nodes are depicted as squares while terminal nodes as circles.Please note that the visual style used for this network is slightly different from that used in Figure 5 of the corresponding publication.