A efficient method to construct BWT index of a given DNA sequence, especially useful for gigantic and high similar genome. DeBWT has good scalability to construct BWT in parallel computing. It is well-suited to run on multiple core servers or clusters to construct the BWT of large collections of genome sequences.