The results of Table III suggest that the OVERLAPPING CORRELATION-CLUSTERING
algorithms, find a clustering that
is better in agreement with the SCOP taxonomy than are the
clusterings found by SCPS. However, while allowing overlaps
is beneficial, we do not observe a significant improvement
as the node-specific constraint p is increased. Moreover, we
observe that only a small number of proteins are assigned
to multiple clusters. We conjecture that this is due to the