Identifying Dynamic Protein Complexes Based on Gene Expression Profiles and PPI Networks

Algorithm 1

The description of subroutine for detecting possible protein-complex cores.

Subroutine DetectingPPC

(**Detecting possible protein-complex cores**)

Input: 𝐺(𝑉,𝐸): a PPI network; Ge: gene expression data;

  𝑇_𝑔: the threshold of gene expression; 𝑇_𝑑: the threshold of density.

Output: PPC: possible protein-complex cores

(1) for each protein 𝑣∈𝑉 and 𝑣∈Ge  do

    if min(Ge(𝑣)) ≥𝑇_𝑔 then 𝑃_𝐴←𝑣; tag(𝑣) = 0;

(2) for each edge (𝑢,𝑣)∈𝐸  do

    Comptute its edge clustering coefficient ECC(𝑢,𝑣)

(3) 𝑄← sorting proteins in 𝑃_𝐴 by multiple keys 𝑁^𝑃𝐴_𝑣,𝑁^𝑅_𝑣 in descendant order

(4) While  𝑄≠𝚽  do

      𝑣←𝑄;

      if tag(𝑣) = 1 then continue;

      else 𝐶={𝑣};

      For each node 𝑢∈𝑁_𝑣  do

          if tag(𝑢) = 0 then 𝐶=𝐶∪{𝑢};

      While  𝑑𝑒𝑛(𝐶)<𝑇_𝑑  do

       Remove the nodes with minimum ECC(𝑥,𝑣) from 𝐶;

      For each node 𝑢∈𝐶  do

         tag(𝑢) = 1

      PPC = PPC ∪  {𝐶}

(5) Output PPC