IgGraph

What is IgGraph?

IgGraph will provide the germline V, D, and J gene-segment usage from antibody (immunosequencing) reads. A colored de Bruijn graph is used to determine the generating V, D, and/or J gene-segment used for every read.

Input:

• FASTA file of unique immunoglobulin sequencing reads
• FASTA files of V, D, J germline gene-segments

Output:

Tab delimited file with the following

• One row for each input read
• The ID of each identified V, D, J gene-segment
• Identified CDR3 sequence

Dependencies

IgGraph has been tested on many different versions of Ubuntu, Mac OS X, and should work on most unix-based OSs. If building from source the dependencies listed are contained as subtrees in the git repository, so there is no need to install them separately. Instructions for compilation are located below.

• gcc with C++11 support (GCC 4.7 and later)
• SeqAn 2.0+
• TCLAP
• cereal
• catch

Compilation

Clone the repository, all dependencies are contained as subtrees, so nothing special needs to be done. Alternatively, you can also download the tar/zip and build from those.

$ git clone https://github.com/sbonisso/ig_graph.git

Compile by running make in the top-level directory.

$ cd ig_graph
$ make

To create tests, additionally run make test.

$ make test
$ ./unit_test

Usage

The iggraph executable can be run directly, or for simplicity, you can use the two run_iggraph_{hsap,mmus}.sh scripts, from inside the ig_graph directory, to run for either human or mouse.

These scripts will call iggraph with useful parameter values and germline sequences, and only need the input file, output file, and optionally, the number of threads specified as input, for example:

$ ./run_iggraph_hsap.sh --read_file igseq_reads.fa \
      --output_file igseq_vdj.tsv \
      --num_thread 4

These scripts use germline references included in the repository, but any reference gene-segments can be passed as input in the form of FASTA files.

The options and flags can be seen when running the built-in help:

$ ./iggraph -h

Speed

In addition to being accurate, IgGraph is very fast, faster than the few other antibody classification tools available for download.

Citation

If the software is used in a publication, please cite:

Bonissone, S.R. and Pevzner, P.A. (2015). Immunoglobulin classification using the colored antibody graph, Research in Computational Molecular Biology, Lecture Notes in Computer Science, vol. 9029, Springer International Publishing, pp. 44-59. [DOI]