Skip to content

Latest commit

 

History

History
89 lines (58 loc) · 3.5 KB

manual.md

File metadata and controls

89 lines (58 loc) · 3.5 KB

Installation

Requirements

  • 64 bit Linux OS
  • CMake version 3.1 or above
  • C++ compiler with C++14 support (GCC 5.0+)
  • GNU make
  • zlib

Downloading and compiling from source code

To build from source code run the following commands from code directory.

    cmake .
    make

Binary file will be stored in bin subdirectory. We suggest that you add bin directory to your PATH cariable or move the binary file to a directory that is already in your PATH cariable. In the examples below we assume that you did so. Otherwise replace jumboDB with full path.

Running de Bruijn graph construction

Input

Input reads can be in fasta or fastq format and can be compressed using gzip. Make dure that file name extensions correspond to contents of files since they are used to determine file format. E.g. uncompressed fasta reads are expected to be stored in files with name extensions ".fasta" or ".fa" while compressed fastq reads are expected to be stored in files with name extensions ".fq.gz" or "fastq.gz".

Command line

To run de Bruijn graph construction use the following command line

    jumboDB [options] -o <output_dir> --reads <reads_file> [--reads <reads_file2> ...] -k <int>

Basic options

-o <file_name> (or --output-dir <file_name>) Name of output folder. Resulting graph will be stored there.

-k <int> Value of k (vertex size) to be used for de Bruijn graph construction. k should be odd (otherwise k + 1 is used instead).

`--reads <file_name> Name of file that contains reads in fasta or fastq format. This option can be used any number of times in the same command line resulting in collecting reads from multiple files.

-h (or --help) Print help message.

Advanced options

-t <int> (or --threads <int>) Number of threads. The default value is 16.

-w <int> (or --window <int>) The window size to be used for sparse de Bruijn graph construction. The default value is 2000. Note that all reads of length less than k + w are ignored during graph construction.

--compress Compress all homolopymers in reads.

`--coverage Calculate edge coverage of edges in the constructed de Bruijn graph. The coverage is then printed as KC option in gfa files and is printed in edge labels in dot files.

Output of de Bruijn graph construction

All output files are stored in <output_dir> `, which is set by the user.

  • <output_dir>/graph.fasta sequences of all edges of de Bruijn graph in fasta format
  • <output_dir>/graph.gfa sequences of all edges of de Bruijn graph in gfa format. When --coverage option is used, k-mer coverage of each edge (total number of occurences of all k-mers from an edge) is stored as additional parameter KC. Note that gfa format is ill suited for de Bruijn graph storage since it does not represent some vertices such as vertices with indegree 2 and out degree 0.
  • <output_dir>/graph.dot sequences of all edges of de Bruijn graph
  • <output_dir>/dbg.log log file for the run

Feedback and bug reports

Your comments, bug reports, and suggestions are very welcomed. They will help us to further improve de Bruijn graph construction tool. If you have any troubles running graph construction, please send us dbg.log file from the directory <output_dir>.

You can send your comments and bug reports via e-mail: [email protected].