We choose the PPI data of S. cerevisiae to conduct our experiments, since its PPI and essentiality data are the most complete and reliable. The PPI data is downloaded from DIP database (Salwinski et al., 2004), updated on October 10, 2010, with a format “tab”. There are several types of interactions included in the PPI data, such as
physical association, direct interaction and covalent binding. We extract the uniprotkb IDs of all the proteins involved in
the PPI data and then use the “ID Mapping” function in UniProt (http://www.uniprot.org/) to transform the uniprotkb IDs to protein names we need. After deleting the self-interactions and the repeated interactions, the static PPI network consists of 5093 proteins and 24,743 interactions.