songweizhi / Binning_refiner

Publication

• Song WZ, Thomas T (2017) Binning_refiner: Improving genome bins through the combination of different binning programs. Bioinformatics, 33(12), 1873-1875. PDF
• Weizhi Song ([email protected]), Torsten Thomas([email protected])
• The Centre for Marine Bio-Innovation (CMB), University of New South Wales, Sydney, Australia

Dependencies:

• BioPython

Change Log:

Version 1.4.0 (2019-04-19):

• Binning_refiner can now supports MORE THAN THREE binning programs.

Version 1.3 (2019-01-06):

• Binning_refiner can now be installed with "pip install Binning_refiner".

Version 1.2 (2017-11-30):

• Binning_refiner has been simplified to keep only the core functions, which made it much easier to install and use, hope you enjoy it :)

Important message

1. Binning_refiner can now supports MORE THAN THREE binning programs.

2. Binning_refiner can now be installed with "pip install Binning_refiner".

3. An independent R version of Binning_refiner is available now [Binning_refiner.R]. The R packages seqinr, assertr, tidyr, tools, optparse, googleVis are required for running Binning_refiner.R.

4. In the original version of Binning_refiner, the blast approach (as described in its publication) was used to identify the same contig among input bin sets. As Binning_refiner was designed to refine bins derived from the same set of assemblies and the blast step is time-consuming (especially for big dataset), the same assembly among different bin sets was identified by its ID rather than blastn, which made Binning_refiner much faster to run and more easier to install.

How to install:

• Binning_refiner can be installed with pip now:

    # For installation
    pip install Binning_refiner
    
    # For update
    pip install --upgrade Binning_refiner
  

• Visualisation of the correlations between input bins (figure below) can be enabled with '-plot' option. The R packages optparse and googleVis are needed for this function.

Help information:

    Binning_refiner -h
      
      -i          input bin folder
      -p          output prefix
      -m          minimal size (Kbp) of refined bin, default: 512
      -plot       visualize refinement with Sankey plot
      -x          the width of sankey plot
      -y          the height of sankey plot
      -q          silent progress report

How to run:

1. The input for Binning_refiner v1.4 is a folder hold all input bins derived from the same set of assemblies, with each binning program produced bins stored in a separated sub-folder. Here is an example of the input_bin_folder.

2. All bin files in the same sub-folder must have the same extension.

3. Binning_refiner now compatible with both python2 and python3.

    # example command
    Binning_refiner -i input_bin_folder -p Human -plot
    
    # For R version Binning_refiner
    Rscript path/to/Binning_refiner.R -i input_bin_folder -p Human
   

Output files:

1. All refined bins larger than defined bin size cutoff.

2. The id of the contigs in each refined bin.

3. The size of refined bins and where its contigs come from.

4. Visualisation of binning refinement.