antiberty

What it is and what it does

AntiBERTy is a BERT-style transformer model trained on natural antibody sequences. It provides a pre-trained encoder for antibody protein analysis, allowing you to extract learned representations of antibody sequences without training from scratch. The package wraps the model in a simple Python interface with methods to generate fixed-dimension embeddings, classify antibody species and chain type, fill in masked positions, and score sequence likelihood.

You instantiate an AntiBERTyRunner, then call methods like embed(), classify(), fill_masks(), or pseudo_log_likelihood() on lists of antibody sequences. It depends on torch and transformers for the underlying model and inference. The model is designed for computational biology workflows where you need to work with antibody sequences programmatically—for example, screening variants, analyzing affinity maturation, or featurizing sequences for downstream machine learning.

Use it for:

• Generate fixed-size vector representations of antibody sequences for clustering, similarity search, or downstream ML models.
• Classify antibody sequences by species origin and chain type automatically.
• Predict missing or masked residues in antibody sequences based on learned patterns.
• Score the likelihood of antibody sequences to identify natural vs. synthetic variants.
• Analyze attention patterns across antibody sequences to understand model reasoning.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

AntiBERTy is a transformer language model pre-trained on antibody sequences that generates embeddings, classifies antibody species and chain types, predicts masked residues, and scores sequence likelihood.

Yes, if you work with antibody sequences and need a pre-trained transformer encoder—the model is specialized and the install is frictionless. No, if you need active maintenance or support; the package is abandoned and has no license metadata. Consider it a research artifact: useful for one-off analysis or as a feature extractor in a larger pipeline, but not for production systems requiring updates or legal clarity.

Install

antiberty on PyPI

pip install antiberty

uv add antiberty

poetry add antiberty

Installing antiberty

Before you install

Installation is straightforward with low friction via pip. However, the package is marked as abandoned with no updates since July 2023, so expect no maintenance or bug fixes going forward.

License in practice

License status is unclear—no SPDX identifier or raw license text is provided in the package metadata, making it difficult to assess legal obligations before use.

Quickstart

pip install antiberty

from antiberty import AntiBERTyRunner

antiberty = AntiBERTyRunner()
sequences = ["EVQLVQSGPEVKKPGTSVKVSCKASGFTFMSSAVQWVRQARGQRLEWIGWIVIGSGNTNYAQKFQERVTITRDMSTSTAYMELSSLRSEDTAVYYCAAPYCSSISCNDGFDIWGQGTMVTVS"]
embeddings = antiberty.embed(sequences)

Requires torch and transformers; model weights are downloaded on first use.

Verify before relying

• Minimum Python version and specific torch/transformers version requirements not documented in metadata.
• Whether pre-trained model weights are bundled or downloaded at runtime, and total download/storage footprint.
• Current status of the GitHub repository (archived flag is null in metadata).

Package facts

Evidence: antiberty-0.1.3-py3-none-any.whl

Tags

More Artificial Intelligence packages