Estimate average coverage and create Nonpareil curves for metagenomic datasets
Estimate average coverage and create Nonpareil curves for metagenomic datasets.
Nonpareil binary: Nonpareil requires a C++ compiler. It has been tested on 64-bit machines with GCC versions ≥ 4.2.1, running Mac OSX and Read Hat Linux.
Nonpareil utilities: Requires R. No additional libraries are necessary.
Download TAR Ball and unpack it using:
tar zxvf nonpareil.tar.gz
Change directory into the newly created folder, and compile Nonpareil using:
make nonpareil
If you want to make Nonpareil available system-wide, just copy the generated binary into a folder listed in $PATH. For example:
sudo cp nonpareil /usr/local/bin/
Type nonpareil with no arguments to see all the available parameters. To run it with default parameters, simply type:
nonpareil -s reads.fa -b output
Where reads.fa is a FastA file containing the WGS reads and output is the prefix of the files to be generated.
The output files are all in raw text, and can be processed in R to generate Nonpareil curves and estimates. The Nonpareil.R file contains the Nonpareil.curves function. Refer to the embedded documentation for more details. Nonpareil curves with default parameters can be generated in R with:
source('utils/Nonpareil.R');
Nonpareil.curves('output.npo', 50);
Adjust 'utils/Nonpareil.R' and 'output.npo' to point to the actual location of the files. 50 indicates the read overlap used in nonpareil (-L parameter). If you used a different value, change it accordingly.
Luis M. Rodriguez-R & Konstantinos Konstantinidis, at the Kostas lab.