Generate a Reference

If you do not have a reference, you can generate one by using the dragen –build-hash-table command and passing in the location of the reference FASTA file. You can specify a set of parameters when building your hash table (see the DRAGEN Bio-IT Platform User Guide (1000000070494)).

For testing purposes, you can run the example shell script or the one of commands shown in the examples in this guide. For these examples, the FASTA file is assumed to be located in /staging/human/reference/hg19/hg19.fa. Change the path in the script or command line to the correct directory, if needed. You must have change access to /staging/human/reference and its subdirectories.

Run the example shell script as follows:

/opt/edico/examples/build_hash_table.sh

Or, run the dragen command as follows:

mkdir -p /staging/human/reference/hg19/hg19.fa.k_21.f_16.m_149

cd /staging/human/reference/hg19/hg19.fa.k_21.f_16.m_149 \

dragen --build-hash-table true --ht-reference /staging/human/reference/hg19/hg19.fa \
--output-dir /staging/human/reference/hg19/hg19.fa.k_21.f_16.m_149 \
--ht-alt-liftover /opt/edico/liftover/hg19_alt_liftover.sam

If you generate a hash table without including the --ht-alt-liftover option, an error similar to the following may occur (depending on the .fa reference file used):

ERROR: Detected hg19 alternate contigs in reference at:
/staging/hg19fa/hg19.fa
DRAGEN map quality is significantly improved by building a reference with a liftover file to enable ALT aware mapping. Use the --ht-alt-lifeover option to specify a liftover file.
You may ignore this error and continue using your existing reference by adding --ht-alt-aware-validate=false to your command line. However, DRAGEN map quality will be significantly affected.
Generate the hash table with either the --ht-alt-liftover or the –ht alt-aware-validate=false option to avoid the error listed above.

The dragen --build-hash-table command is multithreaded and defaults to eight threads. This command takes approximately 15 minutes to run. You can use the --ht-num-threads option with a value up to 32 (depending on the number of threads your server supports) to reduce the run time.

The hash table directory name lists key default option values that were used during the hash table build. Illumina recommends following this best practice when you generate your own hash tables and change the directory name accordingly.

If you enabled the CNV function, generating a hash table takes ~2 hours.