[Version 1.0.0] Initial Release of CLEF - Code for Contrastive Training and Effector Prediction

[Version 1.0.0] - [2025/1/2]

This is the initial release of the code associated with the manuscript titled "Contrastive-learning of Language Embedding and Biological Features for Cross Modality Encoding and Effector Prediction" . This release provides the implementation of the model for contrastive training of protein language model embeddings and other protein features and predicting bacterial type III/VI/VI secretion effectors. It includes corresponding dependencies and dataset for the study.

Key Features:

• Model: CLEF - a contrastive learning framework for effector prediction.
• Data: Dataset of 10,831 proteins sequences used for unsupervised contrastive training, several datasets for supervised effector prediction training.
• Environment: Python 3.8+, PyTorch 2.0.0.
• Instructions: Please refer to the README file for detailed instructions on setting up and running the code.

Changelog:

• v1.0.0: Initial release with core functionality for contrastive training and bacterial effector prediction.

For any questions or issues, please refer to the repository's Issues section or contact Yue Peng at [756028108@qq.com].