Fig. 2.

Capture of protein complexes in a human DNA repair network. a Hierarchical biclustering on TopS values of the 17 baits and 535 preys in a human DNA repair protein interaction network. Proteins identified in complexes in the DNA repair dataset were hierarchically clustered using normalized TopS values as input using ClustVis⁴⁶. a Scaling method was used to divide the values by standard deviation so that each row had a variance equal to one. Rows were centered and unit variance scaling was applied to rows. Rows were clustered using correlation distance and average linkage. Columns were clustered using correlation distance and Ward linkage. b Bait-specific protein complex enrichment. Two complexes are shown that exhibited high association with specific baits. Red color corresponds to high TopS scores. Subunits of the polycomb complex exhibited high scores with the CBX8 bait, and components of the BRAFT complex exhibited high scores with the RPA3 bait. c Interaction network between baits and known protein complexes identified in the dataset. The network was constructed using the Cytoscape platform²³. Large nodes correspond to a larger number of links. TopS topological scoring