An Improved Binary Differential Evolution Algorithm to Infer Tumor Phylogenetic Trees

Algorithm 1

An improved binary differential evolution algorithm to infer tumor phylogenetic trees (BDEP).

Require: The copy number profiles (object nodes set ).

The max generation .

The number of individuals (population scale) .

Ensure: The tumor Steiner tree with the shortest length.

  Generate candidate Steiner node according to the copy number profiles, construct a complete directed graph .

  Set the generation number , initialize a population of individuals with

where is a binary variable.

  while stopping criterion is not satisfied  do

   Mutation step

    for   to   do

   Generate a mutant individual from the target individual and two different individuals

 , .

   for   to   do

   end  for

    end for

   Crossover step

  for   to   do

  Search the of individual , the best neighbor of is =

     Update trial individual to

     for   to   do

   end for

   end for

    Selection step

  for   to   do

   Evaluate the trial individual

   if    then

   ,

  end if

   end for

   Update the generation count  

  end while

  return optimal tumor Steiner tree