How to create binary sequence files in BNIB format

If you are setting up your own QuickLoad data source, or you just wish to partially load sequence files, you may wish to create BNIB files.

The IGB package contains a java class called ''com.affymetrix.igb.parsers.!NibbleResiduesParser'' that contains methods for reading and writing files in the BNIB format.

It contains a ''main()'' method that can read a FASTA format file and output a BNIB file.

To convert a FASTA file to BNIB, run the following command:

where

seq_name is the name of the sequence, e.g., like "chr1"

in_file is the input sequence file in FASTA format

out_file is the name of the BNIB output file

seq_version is the name of the genome version, like "R_norvegicus_Jun_2003".

The seq_version name should match the version name you use in your QuickLoad directory structure.

You will need to create one BNIB file for each chromosome, and each file should be named with names like chr1.bnib, using the same prefix as the names of the chromosomes in the QuickLoad mod_chromInfo.txt file.

For example, assume you are working with rat genome "rn3" and you have FASTA files for each chromosome. You should have one FASTA file for each chromosome, with only one sequence per file. You can run a command like this to convert the chr1.fa file for chromosome 1 into chr1.bnib:

On a Unix system, you would type this command into a terminal window, making sure that
the file igb_exe.jar was in the current working directory.

You can obtain igb_exe.jar from the downloads section of the IGB/Genoviz site.