MiAIRR-to-NCBI Specification#

Outline of INSDC reporting procedure#

TODO: Outline the reporting procedure for data sets 1-4

In terms of standard compliance it is currently REQUIRED [1] to deposit information for MiAIRR data sets 5 and 6 in general-purpose sequence repositories for which an AIRR-accepted specification on information mapping MUST exist. However, users should note that in the future additional AIRR-sanctioned mechanisms for data deposition will become available as specified by the AIRR Common Repository Working Group. The mapping of data items in MiAIRR data sets 5 and 6 differs substantially in size and structure and therefore requires distinct reporting procedures:

  • Set 5: This is free text information describing the work flow, tools and parameters of the sequence read processing. It is REQUIRED that this information is deposited as a freely available document, permanently linked via a DOI. Note that is currently neither a specific format for this document nor a recommended service provider for obtaining the DOI.

  • Set 6: This is specified to contain the consensus sequence and the following information obtained from the initial analysis: V, D and J segment, C region and IMGT-JUNCTION [2] [LIGMDB_V12]. These will be deposited in a general-purpose INSDC repository, using the record structure described below.

INSDC records were originally designed to hold individual Sanger sequences. Therefore each record will contain a header with information largely identical between all records in an AIRR sequencing study. Records can be concatenated for uploading.

The INSDC feature table (FT) [INSDC_FT] is a sequence annotation standard used within the INSDC records and assigns information to specified positions on the reported sequence string. In regard to the correct location of the provided annotation, it should especially be noted that some V(D)J inference tools will return coordinates referring to the reference instead of the query sequence. As the sequence submitted in a record MUST be identical to the query sequence, the positions provided by the V(D)J inference tool MUST, if necessary, be translated back onto the query sequence. In case the start and/or end of a feature cannot be reliably determined or is not present in the reported sequence [3], open intervals CAN be used for reporting. However, open intervals MUST NOT be used to deliberately obfuscate known positions.

In addition to the required information specified in Table_1, users CAN use all valid FT keys/qualifiers to provide further annotation for the reported sequences. However, a record MUST still be compliant with this specification, if such OPTIONAL information would be removed, meaning that it is FORBIDDEN to move REQUIRED information into OPTIONAL keys/qualifiers. In addition, users MUST NOT use keys/qualifiers that could create ambiguity with the keys/qualifiers specified here.


FT key

FT qualifier

FT value

REQUIRED (if used by original study)

V segment



see [Feature table]


D segment



see [Feature table]

yes; if IGH, TRB or TRD sequence

J segment



see [Feature table]


C region



see [Feature table]







Table 1: Summary of the mapping of mandatory AIRR MiniStd data set 6 elements to the INSDC feature table (FT). Note that the overall record will contain additional information, such as cross-references linking the deposited sequence reads and metadata.

Element mapping#

The broad strategy of element mapping to the various repositories is depicted in Table_2.

MiAIRR data set / subset

target repository

1 / study


1 / subject

1 / diagnosis & treatment

2 / sample


3 / processing (cells)

3 / processing (nucleic acids)


4 / raw sequences

5 / processing (data)

user-defined DOI

6 / Processed sequences & annotations


Table 2: Summary of the mapping of MiAIRR data sets to the various repositories

Mapping of data sets 1-4 to BioProject/BioSample/SRA#

TODO: Include item-by-item mapping [NCBI_NBK47528]

Mapping of data set 5 to a user-defined repository#

While several mandatory item have been defined in this data set, there is currently no mapping as the reporting procedure is implemented as a free text document. AIRR RECOMMENDS to use Zenodo for deposition of these documents, as it is hosted by CERN and supports versioned DOIs (termed “concept” DOI). Users SHOULD use the existing AIRR tag when submitting documents to increase the visiblity of their study.

Mapping of data set 6 to INSDC#

Users should note that while the FT is standardized, the overall sequence record structure diverges between the three INSDC repositories. The following section refers to items at or above the hierarchy level of the FT using the GenBank specification [GENBANK_FF], the corresponding designations of ENA [ENA_MANUAL] are provided in parenthesis [11].

Record header#

The header MUST contain all of the following elements:

  • REQUIRED: header structure as specified by the respective INSDC repository [ENA_MANUAL] [GENBANK_FF] [GENBANK_SR].

  • FORBIDDEN: The DEFINITION entry will be autopopulated by information provided in the FT part (misc_feature, /note).

  • REQUIRED: identifier of the associated SRA record (MiAIRR data set 4) as DBLINK (ENA: DR line). Note that it is not possible to refer to individual raw reads, only the full SRA collections can be linked.

  • REQUIRED: in the KEYWORDS field (ENA: KW line):

    • the term “TLS”

    • the term “Targeted Locus Study”

    • the term “AIRR”

    • the term “MiAIRR:<x>.<y>” with <x> and <y> indicating the used version and subversion of the MiAIRR standard.

  • REQUIRED: DOI of the associated free-text record containing the information on data processing (MiAIRR data set 5) as REMARK within a REFERENCE [4] (ENA: RX line).

  • OPTIONAL: The use of structured comments is currently evalutated for use in future versions of the MiAIRR standard.

Feature table#

The feature table, indicated by FEATURES (ENA: RX line), MUST or SHOULD contain the following keys/qualifiers:

General sequence information#
  • REQUIRED: key source containing the following qualifiers:

    • REQUIRED: qualifier /organism (required by [INSDC_FT]).

    • REQUIRED: qualifier /mol_type (required by [INSDC_FT]).

    • REQUIRED: qualifier /citation pointing to the reference in the header (REFERENCE, ENA: RN line) that links to the data set 5 document.

    • REQUIRED: qualifier /rearranged [5].

    • REQUIRED: qualifier /note containing the AIRR_READ_COUNT keyword to indicate the read number used for the consensus. The criteria for selecting these reads and the procedure used to build the consensus SHOULD be reported as part of data set 5.

    • OPTIONAL: qualifier /note containing the AIRR_INDEX_CELL keyword for single-cell experiments. The value of the keyword SHOULD only contain alpha-numeric characters and MUST be identical for sequences derived from the same cell of origin.

    • RECOMMENDED: qualifiers /assembly_gap and /linkage_evidence to annotate non-overlapping paired-end sequences.

    • RECOMMENDED: qualifier /strain, if /organism is “Mus musculus”.

Note that additional qualifiers might be REQUIRED by GenBank to harmonize the GenBank record with the BioSample referenced by it in the header. A list of known BioSample keyword and GenBank qualifiers that MUST contain the same information can be found below. Whether (and in which direction) the existence of a keyword/qualifiers triggers a requirement in the corresponding record is currently unknown. Please report any undocumented requirements surfacing during submission to the MiAIRR team.

BioSample keyword

GenBank FT qualifier

cell type








Segment and region annotation#

The following keys MUST be used for annotation according to their FT definition, if the respective item has been reported by the original study:

  • REQUIRED: key V_region. Note that this key MUST NOT be used to annotate V segment leader sequence [6] [7].

  • REQUIRED: key misc_feature with coordinates identical to those given in V_region. This key MUST contain a /note qualifier that contains a string as value, which describes the general type of variable region described by the record. The string MUST match the regular expression

    /^(immunoglobulin (heavy|light)|T cell receptor (alpha|beta|gamma|delta)) chain variable region$/

    This string will be used as record heading upon import into Genbank. Note that while this behavior of Genbank is undocumented, the procedure has been approved by NCBI.

  • REQUIRED: key V_segment, both coordinates MUST be within V_region. Note that this key MUST NOT be used to annotate V segment leader sequence [6] [7].

  • REQUIRED: key D_segment, both coordinates MUST be within V_region. This key is only REQUIRED for sequences of applicable loci (IGH, TRB, TRD [8]). In the rare case of rearrangements using two D segments, this key SHOULD occur twice, but the coordinates of both keys MUST NOT overlap.

  • REQUIRED: key J_segment, both coordinates MUST be within V_region.

  • REQUIRED: key C_region, both coordinates MUST NOT overlap with V_region. If the region can be unambiguously identified, the respective official gene symbol MUST be reported using the /gene qualifier. If only the isotype (e.g. IgG) but not the subclass (e.g. IgG1) can be identified, a truncated gene symbol (e.g. IGHG instead of IGHG1) SHOULD be reported instead [9].

Each [VDJ]_segment key MUST or SHOULD contain the following qualifiers:

  • REQUIRED: qualifier /gene, containing the designation of the inferred segment, according to the database in the first /db_xref entry. This qualifier MUST NOT contain any allele information.

  • RECOMMENDED: qualifier /allele, containing the designation of the inferred allele, according to the database in the first /db_xref entry. Note that while INSDC does not specify any format for this qualifier, AIRR compliance REQUIRES that this field only contains the allele string, i.e. without the gene name or separator characters.

  • REQUIRED: qualifier /db_xref, linking to the reference record of the inferred segment in a germline database [INSDC_XREF]. This qualifier can be present multiple times, however only the first entry is mandatory and MUST link to the database used for the segment designation given with /gene and (if present) /allele.

    Note on referencing IMGT databases: There are two IMGT database available in the controlled vocabulary [INSDC_XREF]:

    • IMGT/GENE-DB: This is the genome database, which requires that a reference sequence has been mapped to genomic DNA. When using this database as reference, note that you can only refer to the gene symbol not the allele. In the case of ambiguous allele calls (see below) this means that you MUST NOT annotate any /allele at all. Nevertheless, this SHOULD be the default database for applications using IMGT as reference, as the sequence for each gene/allele is unique.

    • IMGT/LIGM: This database collects sequences described in INSDC databases (GenBank/ENA/DDBJ). As it might contain multiple entries representing a given gene/allele, it is NOT RECOMMENDED to use it unless that inference gene/allele is only present in IMGT/LIGM and not in IMGT/GENE-DB.

  • RECOMMENDED: /inference to indicate the tool used for segment inference. The description string SHOULD use COORDINATES as category and aligment as type [INSDC_FT].

Annotation of sequences producing multiple hits with identical scores is problematic and is ultimately at the discretion of the depositing researcher. However, the algorithms used for tie-breaking SHOULD be documented in data set 5. In addition, the following procedures MUST be followed:

  • Certain gene, ambiguous allele: If multiple alleles of the same gene match to the sequence, the /allele qualifier MUST NOT be used. As the REQUIRED /db_xref qualifier will ofter refer to a specific allele, all equal hits SHOULD be annoted via this qualifier (which can be use multiple times). Also see the note on the limitations of the IMGT/GENE-DB reference database above.

  • Ambiguous gene: Pick one, annotate using the qualifiers as noted for ambiguous allele.

JUNCTION annotation#

INSDC does currently not define a key to annotate JUNCTION [10]. Therefore the following procedure MUST be used:

  • REQUIRED: key CDS, indicating the positions of

    1. the first bp of the first AA of JUNCTION

    2. the last bp of the last AA of JUNCTION as determined by the utilized V(D)J inference tool.

    Open coordinates MUST be used for both coordinates to allow for automated creation of the /translated qualifier providing the peptide sequence. Further note that a non-productive JUNCTION can have a length not divisible by three. This key contains the following qualifiers:

    • REQUIRED: qualifier /codon_start with the assigned value “1”.

    • REQUIRED: qualifier /function with the assigned value “JUNCTION”.

    • REQUIRED: qualifier /product with an assigned value matching the regular expression

      /^(immunoglobulin (heavy|light)|T cell receptor (alpha|beta|gamma|delta)) chain junction region$/

      The variable region referred to in the string MUST be the same as the one given in the misc_feature key.

    • RECOMMENDED: qualifier /inference, indicating the tool used for positional inference. The description string SHOULD use COORDINATES as category and protein motif as type [INSDC_FT].

    • FORBIDDEN: qualifier /translated, which will be automatically added by Genbank.

    Note that the complete CDS key will be removed by Genbank if the translation contains stop codons or to many “N” (exact number unknown). As such a record will lack a central piece of REQUIRED information it is RECOMMENDED that submitters either

    • remove the complete record or

    • replace the CDS with a misc_feature key while at the same time removing the /codon_start and /product qualifiers

    upfront, as described in the submission manual. If the submitter chooses the replacement option, it has to be ensured that the annotated coordinates are actually valid and not affect by the frame- shift.

Record body#

The record body starts after ORIGIN (ENA: SQ line) and MUST contain:

  • the consensus sequence



IMGT-ONTOLOGY definitions. <http://www.imgt.org/ligmdb/label#JUNCTION>

[INSDC_FT] (1,2,3,4,5)

The DDBJ/ENA/GenBank Feature Table Definition. <http://www.insdc.org/documents/feature-table>

[ENA_MANUAL] (1,2)

European Nucleotide Archive Annotated/Assembled Sequences User Manual. <http://ftp.ebi.ac.uk/pub/databases/ena/sequence/release/doc/usrman.txt>

[GENBANK_FF] (1,2)

GenBank Flat File Format. <https://ftp.ncbi.nih.gov/genbank/gbrel.txt>

[INSDC_XREF] (1,2)

Controlled vocabulary for /db_xref qualifier. <http://www.insdc.org/documents/dbxref-qualifier-vocabulary>



Example record (GenBank format)#

LOCUS       AB123456                 420 bp    mRNA    linear   EST 01-JAN-2015
DEFINITION  TLS: Mus musculus immunoglobulin heavy chain variable region,
VERSION     AB123456.7
KEYWORDS    TLS; Targeted Locus Study; AIRR; MiAIRR:1.0.
SOURCE      Mus musculus
  ORGANISM  Mus musculus
            Eukaryota; Metazoa; Chordata; Craniata; Vertebrata;
            Euteleostomi; Mammalia; Eutheria; Euarchontoglires; Glires;
            Rodentia; Sciurognathi; Muroidea; Muridae; Murinae; Mus.
REFERENCE   1  (bases 1 to 420)
  AUTHORS   Stibbons,P.
  TITLE     Section 5 information for experiment FOO1
  JOURNAL   published (01-JAN-2000) on Zenodo
  REMARK    DOI:10.1000/0000-12345678
REFERENCE   2  (bases 1 to 420)
  AUTHORS   Stibbons,P.
  TITLE     Direct Submission
  JOURNAL   Submitted (01-JAN-2000) Center for Transcendental Immunology,
            Unseen University, Ankh-Morpork, 12345, DISCWORLD
DBLINK      BioProject: PRJNA000001
            BioSample: SAMN000001
            Sequence Read Archive: SRR0000001
FEATURES             Location/Qualifiers
     source          1..420
                     /organism="Mus musculus"
     V_region        1..324
     misc_feature    1..324
                     /note="immunoglobulin heavy chain variable region"
     V_segment       1..257
     D_segment       266..272
     J_segment       291..324
     CDS             <258..>290
                     /product="immunoglobulin heavy chain junction region"
                     /inference="COORDINATES:protein motif:IgBLAST:1.6"
     C_region        325..420
        1 agcctggggc ttcagtgaag atgtcctgca aggcttctgg ctacacattc actgactata
       61 acatacactg ggtgaagcag agccatggaa agagccttga gtggattgca tatattaatc
      121 ctaacaatgg tggttatggc tataacgaca agttcaggga caaggccaca ttgactgtcg
      181 acaggtcatc caacacagcc tacatggggc tccgcagcct gacctctgag gactctgcag
      241 tctattactg tgcaagagcg ggagtttacg acggatatac tatggactac tggggtcaag
      301 gaacctcagt caccgtctcc tcagccaaaa caacagcccc atcggtctat ccactggccc
      361 ctgtgtgtgg aggtacaact ggctcctcgg tgactctagg atgcctggtc aagggcaact


  • MUST / REQUIRED: Indicates that an element or action is necessary to conform to the standard.

  • SHOULD / RECOMMENDED: Indicates that an element or action is considered to be best practice by AIRR, but not necessary to conform to the standard.

  • CAN / OPTIONAL: Indicates that it is at the discretion of the user to use an element or perform an action.

  • MUST NOT / FORBIDDEN: Indicates that an element or action will be in conflict with the standard.


  • AA: amino acid

  • bp: base pair

  • DOI: digital object identifier

  • FT: INSDC Feature Table

  • INSDC: International Nucleotide Sequence Database Collaboration

  • SRA: sequence read archive