Sequence analysis

IgBLAST

Annotate antibody and T cell receptor sequences against germline V, D and J genes.

IgBLAST is NCBI's aligner for immunoglobulin and T cell receptor variable domains. It reports the top germline V, D, J and constant-region matches for each query, the rearrangement summary (chain type, stop codons, V-J frame, productivity), the V-D and D-J junction details with their N/P nucleotides, and the framework and complementarity determining regions under either the IMGT or the Kabat delineation. Nucleotide queries run through igblastn and protein queries through igblastp; a nucleotide run also writes the AIRR Rearrangement TSV and, on request, a clonotype summary of the whole repertoire.

Documentation ↗ Input parameters ↗ Examples ↗ Home page ↗ Paper ↗ Source code ↗ License ↗

No jobs before this; it will run immediately.

Not logged in. Register an account to keep track of your jobs and results, and add presets and bookmarks. Email results to this address when complete:

Ready

Example preset

Choose an example to fill in its settings and input structures.

Input mode

Query

Official documentation ↗

Job name Names the files this run leaves: <job name>.airr.tsv, <job name>.igblast.txt and <job name>.clonotypes.txt.

Query type igblastn aligns a rearranged nucleotide sequence against V, D and J germline genes and can also call the constant region. igblastp aligns a translated V domain against germline V genes only: D, J, C, the junction and the AIRR table do not apply to it. In this form: One field chooses the program, because IgBLAST ships the two as separate executables.

Query sequence or FASTA One bare sequence (line wrapping allowed) or FASTA records with unique headers; every record is annotated independently. Nucleotide queries accept A, C, G, T, U and the IUPAC ambiguity letters; protein queries accept the amino-acid letters. Up to 5000 records and 400,000 characters per run. In this form: IgBLAST's -query takes a file; what you enter here is written to query.fasta inside the isolated run directory. A designed sequence from an earlier ProteinMPNN, LigandMPNN, AbMPNN or AntiFold run can be dropped in whole from the picker above; those write amino acids, so set the query type to protein.

Query file Upload a FASTA file, or a plain-text sequence. The file's text is sent to the compute node and used exactly as if it had been typed above.

Germline databases

Official documentation ↗

Organism IgBLAST's -organism. Chooses the internal annotation data that places the framework and CDR boundaries, so it has to match the germline databases below. Ig is supported for human, mouse, rat, rabbit and rhesus monkey; TCR only for human and mouse.

Receptor type IgBLAST's -ig_seqtype. TCR needs TCR germline databases, which NCBI does not distribute prebuilt: build them from IMGT with bin/edit_imgt_file.pl and bin/makeblastdb, and point IGBLAST_GERMLINE_ROOT at them.

Germline V database IgBLAST's -germline_db_V. Automatic picks NCBI's recommended prebuilt set for the organism above: the AIRR-C set for human, NCBI's own set for mouse and rhesus monkey. Any BLAST database installed under the germline root can be chosen instead.

Germline D database IgBLAST's -germline_db_D, for nucleotide queries only. Light-chain and TCR alpha/gamma rearrangements have no D segment and IgBLAST simply reports none, so leaving this on Automatic is safe for a mixed repertoire.

Germline J database IgBLAST's -germline_db_J, for nucleotide queries only. The auxiliary coding-frame file is keyed to the J genes, so the CDR3 end and FWR4 are only annotated when the two agree.

Constant region database IgBLAST's -c_region_db, for nucleotide queries only. Identifies the isotype when the query runs into the constant region. NCBI distributes a prebuilt set for human only, so Automatic leaves it unset for every other organism. In this form: A query that stops at the end of the J segment, as most V(D)J amplicons do, has no constant region to call and reports an empty c_call.

Annotation

Official documentation ↗

Domain delineation IgBLAST's -domain_system: which numbering convention the framework and complementarity determining regions are cut at.

Show translated alignments IgBLAST's -show_translation. Prints the amino-acid translation above the nucleotide alignment in the readable report. Nucleotide queries only.

Extend V alignment at the 5' end IgBLAST's -extend_align5end. Forces the V alignment out to the start of the query, which recovers the first framework residues of a read beginning inside FR1.

Extend J alignment at the 3' end IgBLAST's -extend_align3end. Forces the J alignment out to the end of the query.

Minimum V length IgBLAST's -min_V_length: shorter V alignments are not reported. Default 9.

Minimum J length IgBLAST's -min_J_length: shorter J alignments are not reported. Default 0, meaning no minimum. Nucleotide queries only.

Minimum D match IgBLAST's -min_D_match: the consecutive nucleotide bases a D gene must match to be called. Raise it to suppress the spurious short D hits a hypermutated junction attracts. Nucleotide queries only.

V mismatch penalty IgBLAST's -V_penalty. Default -1, which tolerates the somatic hypermutation a matured V segment carries.

D mismatch penalty IgBLAST's -D_penalty. Default -2. Nucleotide queries only.

J mismatch penalty IgBLAST's -J_penalty. Default -2. Nucleotide queries only.

Allow V(D)J genes to overlap IgBLAST's -allow_vdj_overlap. Active only when the D and J mismatch penalties are set to -4 and -3; a run that asks for it without them is refused rather than silently ignoring the option.

Output

Official documentation ↗

Write the AIRR rearrangement table Runs IgBLAST with -outfmt 19 and keeps the result as <job name>.airr.tsv: one row per query with the V, D, J and C calls, the junction, every FWR and CDR, and the per-segment alignment coordinates, under the AIRR Rearrangement schema. This is the file offered at the top of the results page. Nucleotide queries only.

Readable report format IgBLAST's -outfmt for the readable report kept as <job name>.igblast.txt. Format 3 is IgBLAST's own default and the one its documented examples use.

V alignments to report IgBLAST's -num_alignments_V: germline V genes shown per query in the readable report. Default 3.

D alignments to report IgBLAST's -num_alignments_D. Default 3. Nucleotide queries only.

J alignments to report IgBLAST's -num_alignments_J. Default 3. Nucleotide queries only.

C alignments to report IgBLAST's -num_alignments_C. Default 2. Only used when a constant region database is selected.

Top clonotypes to report IgBLAST's -num_clonotype, written to <job name>.clonotypes.txt through -clonotype_out. A clonotype groups the queries sharing their V(D)J calls, productivity and CDR3. Set to 0 to skip the clonotype pass. Nucleotide queries only.

BLAST search options

Official documentation ↗

E-value threshold IgBLAST's -evalue. Leave empty for the program's own default: 20 for igblastn, 1 for igblastp.

Word size IgBLAST's -word_size, the length of the best perfect match the word finder seeds on. Leave empty for the program's default; the minimum is 4 for nucleotide queries and 2 for protein.

Gap open cost IgBLAST's -gapopen. Leave empty for the program's default.

Gap extend cost IgBLAST's -gapextend. Leave empty for the program's default.

Query strand IgBLAST's -strand. Default both, which is what lets IgBLAST report a reverse-complemented read and set rev_comp in the AIRR table. Nucleotide queries only.

Threads IgBLAST's -num_threads. Default 4.

Ready

Endpoint POST /api/igblast/

Send JSON or standard form data. A GET request returns this tool's field schema.

API reference for IgBLAST →