HKC: An Algorithm to Predict Protein Complexes in Protein-Protein Interaction Networks

Pseudocode 1

Procedure Scoring.

Input:

 PPI network (an indirect simple graph): 𝐺=(𝑉,𝐸)

 Node score threshold: T₁ and T₂

 Cohesion threshold: T₃

Output:

 The predicted protein clusters: Clusters

Call Scoring

Call ClusterFinding

Call Filtering

// step1: scoring

Procedure Scoring

 for all node n in 𝑉 do

  score(n) = 0

 end for

 for all node n in 𝑉 do

  compute the degree of n

  N = neighborhood (n) // neighborhood returns the

            // direct neighborhood of n (including n)

  H = hkc(N) // hkc returns the highest k-core of N

  𝑘_max = the maximal k value corresponding to H

  N_H = the number of nodes in H

  den(H) = the density of H

  score(𝐻)=𝑁_𝐻∗den(𝐻)∗𝑘_max

  for all node m in H do

   score(𝑚)=max{score(𝐻),score(𝑚)}  // compute the score of node

  end for

 end for

end procedure