MBBC is a useful tool in metagenomic studies. It is a novel composition-based approach to bin environmental shotgun reads, by considering the k-mer frequency in reads and the inferred Markovian property of the unknown species or OTUs (operational taxonomic units).

Tested on 12 simulated and two real experimental datasets, we showed that MBBC reliably binned reads and determined the species number, the genome size, and the k-mer coverage of each species, for both paired-end reads and single-end reads. MBBC also performed better than two state-of-the-art composition-based methods.


Download the Software and manual.


There are only three parameters for MBBC.
  1. Reads files: paired-end or single-end reads
    Paired-end reads should be in following format:

  2. initial species number m (should be larger than the real species number), by default m=10.

  3. Read type: 0 or 1, by default Read type=0;
    0: denotes paired-end reads;
    1: denotes single-end reads;


MBBC provided us the detailed running process and parameter predictions.
  1. MBBC running process, which are shown in the output window.
  2. MBBC prediction file: a file containing predicted species number, genome size, k-mer coverage and relative abundance before and after reads binning
  3. MBBC binned read files: input reads are binned into separate files.
See manual for the details of how to use MBBC.

For any questions, please contact us with emails:
  • Ying Wang: ying2010@knights.ucf.edu
  • Haiyan Hu: haihu@cs.ucf.edu
  • Xiaoman Li: xiaoman@mail.ucf.edu