Antibody Masked Completion (Ab-X-Completion)

This tool completes X (masked residues) in antibody sequences. We first use an antibody MLM model to progressively fill all X positions (retaining Top-K candidates per site and combining multi-site replacements via beam search), then use an antibody GPT model to compute PPL for each completed sequence and re-rank by PPL from low to high, returning the Top-N optimal completion candidates. The input sequence must contain 1–8 X residues.

Two input modes are supported: ① single-chain sequence (VH or VL, no |); ② paired sequence (VH|VL, exactly one |, with heavy and light chains on either side), e.g.: QVQL...|DIQM....

This tool supports only a single sequence input. You may select the species (Human / Camelid) to match the corresponding reference statistics and scoring framework.

1. Antibody Amino Acid Sequence (supports only 1 FASTA entry):

Parsed sequences: 0, Total residues: 0

Species:



Model Performance (Current Version)

Antibody MLM Model Performance:
  PPL = 1.39
  Acc = 0.9162

Antibody Autoregressive (GPT) Model Performance:
  PPL = 1.47
  Acc = 0.8974

Metrics Description

X is treated as a masked residue to be completed. MLM outputs candidate amino acids and their log-probabilities for each X position, and generates multiple complete sequences via beam search. GPT then computes PPL for each completed sequence and outputs the Top-N candidates sorted by PPL. Sequences without X are not accepted for submission.

Last updated: 2026-05-26