---
title: IgBLAST inputs and parameters
description: Sequences, organism, domain system, and alignment options for IgBLAST.
tags:
  - igblast
  - inputs
---

You provide one or more antibody sequences and choose the organism whose germline database IgBLAST should align against. The organism is the parameter that matters most: aligning a mouse antibody against the human database mislabels every germline gene. Everything else tunes how many hits you see and how strict the alignment is.

## Sequences

IgBLAST accepts a list of antibody variable-region **amino-acid** sequences. Submit between one and 50 sequences per job.

⊘ **Nucleotide input is not accepted.** The nucleotide germline databases (V, D and J) are not installed, so a nucleotide sequence is refused rather than analysed. This also means D and J genes are never assigned and the CDR3 is not named — `igblastp` searches a V database only. Use [ANARCI](/docs/anarci) for CDR3 and region boundaries.

- `sequences` (string[]) —
Antibody variable-region amino-acid sequences. One to 50 per job.

- `sequence_names` (string[]) —
Optional names for each sequence.

## Germline database

- `organism` (string) —
Species whose germline database to align against: `human`, `mouse`, `rat`, or `rabbit`.

- `domain_system` (string) —
Numbering / domain system for boundaries: `imgt` or `kabat`.

## Alignment tuning

- `num_alignments_v` (int) —
Number of top V-gene alignments to report (1–20).

- `num_alignments_j` (int) —
Number of top J-gene alignments to report (1–20).

- `evalue_threshold` (float) —
E-value threshold for reporting hits (greater than 0, up to 100).

## From a therapeutic name

The `igblast_annotate` tool also accepts a `therapeutic_name` — name a known antibody and the platform resolves its VH/VL from Thera-SAbDab before annotating. Submit one name per call; for a multi-antibody panel, make separate parallel calls.

## Read next