A reference-free lossless compression algorithm for DNA sequences using a competitive prediction of two classes of weighted models