---
license: cc-by-nc-nd-4.0
---
# PepDoRA: A Modified Peptide-Specific Language Model via Weight-Decomposed Low-Rank Adaptation

In this work, we introduce **PepDoRA**, a SMILES transformer that fine-tunes the state-of-the-art [ChemBERTa-77M-MLM](https://huggingface.co/DeepChem/ChemBERTa-77M-MLM) transformer on modified peptide SMILES via [DoRA](https://nbasyl.github.io/DoRA-project-page/), a novel PEFT method that incorporates weight decomposition. These representations can be leveraged for numerous downstream tasks, including membrane permeability prediction and target binding assessment, for both unmodified and modified peptide sequences.

Here's how to extract PepDoRA embeddings for your input peptide:

```
import torch
from transformers import AutoModel, AutoTokenizer

# Load the model and tokenizer
model_name = "ChatterjeeLab/PepDoRA"
model = AutoModel.from_pretrained(model_name, output_hidden_states=True)
tokenizer = AutoTokenizer.from_pretrained(model_name)

# Input peptide sequence
peptide = "CC(C)C[C@H]1NC(=O)[C@@H](C)NCCCCCCNC(=O)[C@H](CO)NC1=O"

# Tokenize the peptide
inputs = tokenizer(peptide, return_tensors="pt")

# Get the hidden states (embeddings) from the model
with torch.no_grad():
    outputs = model(**inputs)

# Extract the embeddings from the last hidden layer
last_hidden_state = outputs.hidden_states[-1]

# Print the embedding shape (or the embedding itself)
print(last_hidden_state.shape)
```

## Repository Authors

[Leyao Wang](mailto:leyao.wang@vanderbilt.edu), Undergraduate Intern in the Chatterjee Lab <br>
[Pranam Chatterjee](mailto:pranam.chatterjee@duke.edu), Assistant Professor at Duke University 

Reach out to us with any questions!