Hirbin (High-resolution binning) is a python package for binning and sub-binning of metagenomes at the functional level.
It combines supervised functional annotation using e.g. PFAM, TIGRFAM or COGs, with an unsupervised clustering step
where the sequences in each bin are clustered using Uclust.
The four main hirbin programs are:
functionalAnnotation.py
mappingReads.py
clusterBinsToSubbins.py
statisticalAnalysis.py
Hirbin depends on the following python packages:
numpy (1.8.1)
biopython (1.66)
multiprocessing (2.6.2.1)
It is recommended to create a new conda environment using miniconda and install the required packages inside:
conda create -n hirbin numpy biopython
source activate hirbin
Alternatively install the required python packages first by for example typing pip install biopython.
After installing the python dependencies, download the hirbin source code, extract the files and install the hirbin package:
wget https://github.com/cmbio/hirbin/archive/v0.1.tar.gz
pip install v0.1.tar.gz/
After installation type e.g. functionalAnnotation.py -h to see the options for the program.
Depending on which part of Hirbin you want to use you need to install some external programs and add them to the path environmental variable, so that Hirbin can find them.
The external programs needed are:
- hmmer3 (used by
functionalAnnotation.py) - EMBOSS transeq v6.5.7 (used by
functionalAnnotation.py) - USEARCH (uclust) (used by
clusterBinsToSubbins.py) - R (used by
statisticalAnalysis.py) - Bowtie2 (used by mappingReads.py)
- bedtools (used by mappingReads.py)
- samtools (used by mappingReads.py)