---
pretty_name: p-IgGen
library_name: transformers
tags: [science, antibodies, protein,drug discovery]
---

# Model Card for Model ID

Auto-regressive protein language model for paired antibody library generation.

Pretrained on unpaired heavy and light chain sequences from the Observed Antibody Space (OAS), and finetuned on paired sequences.

## Getting Started?

This model was featured in the Hugging Face blog ***How to Train an Antibody Developability Model*** (found here: https://huggingface.co/blog/ginkgo-datapoints/making-antibody-embeddings-and-predictions). Sample code on how to get started using this model can be found in the blog.

## Model Details

### Model Description


- **Funded by:** EPSRC, AstraZeneca
- **License:**  BSD-3-Clause license

### Model Sources

<!-- Provide the basic links for the model. -->
Note: This is a duplicate of Ollie Turnbull's model repo, found here: https://huggingface.co/ollieturnbull/p-IgGen
- **Repository:** https://github.com/OliverT1/p-IgGen
- **Paper:** https://academic.oup.com/bioinformatics/article/40/11/btae659/7888884