Predicting sgRNA Off-Target Effects Using DNA-Language Embedding and Triplet Attention Mechanism

Nhat-Anh Nguyen-Dang, Phuc H. Le, Thanh-Hoang Nguyen-Vo, Binh P. Nguyen

Abstract

CRISPR-Cas9 has transformed genome editing by enabling precise DNA modifications, but off-target effects remain a significant challenge. This study presents a novel deep learning framework that integrates a DNABERT-6 pretrained sequence encoder with an attention-based convolutional neural network and a customized Triplet Attention mechanism to predict sgRNA off-target effects. Our model processes wild-type and mutated sequence pairs to learn intricate sequence relationships, with the Triplet Attention module effectively emphasizing critical regions. Comparative evaluations against 28 traditional machine learning models and several state-of-the-art deep learning methods demonstrate the superiority of the proposed approach, achieving an Area Under the Receiver Operating Characteristic curve of 0.8954 and an Area Under the Precision-Recall Curve of 0.8798 on the test set. Ablation studies confirm the impact of the Triplet Attention mechanism, while robustness assessments over 30 independent trials show consistently high performance. These results highlight the potential of combining DNA-language embeddings with advanced attention mechanisms to improve the accuracy and reliability of off-target prediction, contributing to safer and more effective CRISPR applications.

Installation

git clone https://github.com/mldlproject/sgRNA-TAC.git
cd sgRNA-TAC
python -m venv .venv
.\.venv\Scripts\activate      # source .venv/bin/activate on Linux/macOS
pip install -r requirements.txt

Usage

Training script (includes evaluation & file export)

python train.py --config configs/default.yaml

Data

Download the PKD off-target dataset (wild-type and mutated sequences with Day21-ETP labels) and place it in data/PKD.csv. The current folder structure expects columns WTSequence (WildType), MutatedSequence, and Day21-ETP-binarized. Additional FASTA exports reside under data/Fasta_PKD/. Replace with your own dataset if needed, keeping the column names consistent.

Citation

@article{nguyen2025sgRNA,
  title={Predicting sgRNA Off-Target Effects Using DNA-Language Embedding and Triplet Attention Mechanism},
  author={Nhat-Anh Nguyen-Dang and Phuc H. Le and Thanh-Hoang Nguyen-Vo and Binh P. Nguyen},
  year={2025},
  note={Manuscript is Under-review}
}

Contact

Free to reach out to me via email.