Greedy Search

Scoring function

Consider a set S = ( g1, , g2, ..., gk ) of k genes.
scoringfunction.gif
where di: the statistic of gene i, μ: the mean of statistic of all background genes.

Search algorithm

  1. Finds genes with more than three interactions.
  2. Constructs sub-networks from each seed genes.
  3. Adds the neighboring genes to sub-networks with maximum scores
  4. Repeats step #3 till stopping conditions meet:
    • No more neighboring gene increases the score
    • Max # of genes in a sub-network

Last edited Jan 31, 2013 at 8:33 PM by yongkeecho, version 2

Comments

No comments yet.