Fig. 6.

Topological network and linkage analysis of a yeast chromatin remodeling network. a A TDA network was constructed for the yeast data inputting TopS scores to the Ayasdi platform²⁴. Distance Correlation was used as a metric with two filter functions: Neighborhood lens 1 and Neighborhood lens 2. Resolution 50 and gain 6x eq. were used. Proteins are colored based on the rows per node. Color bar: red: high values, blue: low values. Node size is proportional with the number of proteins in the node. The location of protein complexes subunits within the topological network are drawn on top of the TDA output. b Hierarchical biclustering using TopS values. Components of the INO80, SWI/SNF, and RSC complexes were clustered using PermutMatrix⁴⁷. Rows and columns were clustered using Pearson as the distance and Ward linkage as the method. Yellow corresponds to high TopS values and gray indicates negative TopS values. Proteins that are not present in the purifications are shown in black. c, d Comparison of TopS values with crosslinking data. Each plot shows the direct interactions between subunits of c INO80 and d SWI/SNF. A red bidirectional edge represents the reported crosslinking interactions with a high TopS value. Black edges represent the reported crosslinking interactions with lower TopS values. TDA topological data analysis, TopS topological scoring