A Fast Cluster Motif Finding Algorithm for ChIP-Seq Data Sets

Algorithm 1

Input: a test set and a control set .

Output: the set of motifs

()  // the set of motifs

()  // the set of emerging substrings

()  // the set of the qualified neighborhood instances

()  // the set of PWMs

()  // the set of intra-motif distributions

() For   to   do

()    For  each -mer of substrings:   do

()       if  (, ) ≥  && (, , ) ≥   then

()         add to

()    For  each -mer of :   do

()      For  each to   do

()      calculate -score of each neighborhood instance

()      if  z() > 1.643  then

()       Add to

()     use and to construct and

()   add to set and add to set

()   For  each of   do

()     if  sim(, ) ≥ 0.75 ()  then

()     cluster with and delete from .

()     if  FDR() > 0.2 then

()    delete from

()   use and corresponding to compute IC.

()   add formed by of top 50 IC score to

()   return