Generating MiAIRR compliant GenBank/TLS submissions


The MiAIRR standard (minimal information about adaptive immune receptor repertoires) is a minimal reporting standard for experiments using sequencing-based technologies to study adaptive immune receptors (T and B cell receptors). The current version (1.0) of the standard was published in Rubelt et al, 2017 and accepted by the general assembly at the annual AIRR Community meeting in December 2017.

MiAIRR recommends submission of raw read data to the Sequence Read Archive (SRA) and submission of processed and annotated data to the Targeted Locus Study (TLS) section of GenBank.

This example will cover generation of files for submission to TLS starting from Change-O formatted data. For complete details of the required and optional elements of the TLS submission see the AIRR Standards documentation site.

Example Data

We have hosted a small example data set resulting from the Roche 454 example workflow described in the pRESTO documentation. The files can be downloded from here:

Change-O Example Files

The following examples use the database file and S43_template.sbt file provided in the example bundle, which has already undergone the IgBLAST annotation, parsing, and filtering operations.

Generating files for submission


Cleaning up C region annotations

First, we need to rename the C primer annotations to valid constant region gene names so they are recognized by remote databases. We’ll use the update subcommand of ParseDb to rename the values in the CPRIMER column of the example file and save that file as update -d -o \
    -f CPRIMER \

Creating ASN files

ASN submission files are generated using the genbank subcommand of ConvertDb as follows: genbank -d \
    --product "immunoglobulin heavy chain" \
    --inf "IgBLAST:1.7.0" \
    --organism "Homo sapiens" \
    --tissue "Peripheral blood" \
    --cell-type "B cell" \
    --isolate S43 \
    --cf CPRIMER \
    --nf DUPCOUNT \
    --asis-id \
    --asn \
    --sbt S43_template.sbt \
    --outdir S43_TLS

The resulting output in the S43_TLS folder will include a number of files. The Sequin file S43_update_genbank.sqn is the file that will be used for submission and the GenBank record file S43_update_genbank.gbf is similar to what the submission will look like once it has been accepted by GenBank.

The command above manually specifies several required and optional annotations. Alternatively, sample information (organism, sex, isolate, tissue_type, cell_type) can be specified in a separate yaml file and provided via the -y argument. Additional harmonized BioSample attributes, which are not convered by the existing commandline arguments, may be provided in the yaml file. Note, the yaml file adds only sample features, so it cannot be used to specify source features (--product, --mol, --inf and --db arguments), parsing arguments, or run parameters (–label`, --exec, etc). Features specified in the yaml file will override equivalent features specified through the corresponding commandline arguments.


The example shown above automatically runs tbl2asn, because the --asn argument was specified. ConvertDb can be run without running tbl2asn, which will generate only the feature table (S43_update_genbank.tbl) and fasta (S43_update_genbank.fsa) files required to run tbl2asn manually via the command:

tbl2asn -p . -a s -V vb -t S43_template.sbt


There is a known issue with the --asn argument. In some environments, for reasons that are presently unknown, tbl2asn may fail to recongizing the input fasta file and report an error stating Unable to read any FASTA records. Running tbl2asn manually should resolve the issue.

Submitting to GenBank/TLS using SequinMacroSend

After generating the .sqn files, you can submit them as MiAIRR compliant GenBank/TLS records using GenBank’s SequinMacroSend service.

When submitting, simply add the keyword AIRR to the subject line in the submission system and it will be routed accordingly.


Currently, the SequinMacroSend system cannot accept files over 512MB in size. For submissions over the size limit, you must split them into smaller files and note in the submission comments that they are a part of a split submission. Note, the .sqn files used for submission are usually about 30 times the size of the original tab-delimited Change-O file. See the split subcommand of ParseDb for one approach to logically dividing large submissions.