Characterization of Protein Complexes and Subcomplexes in Protein-Protein Interaction Databases

Figure 4

Number of size-2 nested groups (one maximal complex and one subcomplex) for a given Jaccard index. The red line represents the whole subset while the blue dots represent just the number of cases coming from more than one paper or database. High Jaccard seems to correlate with information coming from multiple sources, suggesting that high Jaccard in nested groups may come from data integration problems.