J05-3002 have developed a bottomup local multisequence alignment algorithm that uses words and phrases as
hide detail