3.3. Information of dynamic PPI network
In this section, we calculate the number of active proteins and
interactions included in each temporal network. The distribution
of the number of active proteins that appear at a certain number of
time points is also discussed. As shown in Table 2, the number of
proteins and interactions included differ among the 12 temporal
networks. Although each temporal network only contains a
fraction of the total proteins in the static PPI network, the total
number of proteins in the dynamic network accounts for more
than 86% of the total proteins. A fraction of the proteins is also not
included in the dynamic PPI network because those proteins either
do not have gene expression data or are not active at any time
point. In addition, the time complexity of our method is acceptable
because the number of proteins and interactions involved in each
temporal network are not that high. Then, we count the number of
active proteins appear in a certain number of temporal networks
and the result (Fig. 2). Most proteins appear in not more than six
temporal networks and only four proteins are included in all of the
12 temporal networks. This result is caused by the high active
threshold.