CLP

Setup

1. Clone the repository
2. Create and activate a virtual environment
3. Install the required dependencies:

   pip install -r requirements.txt

4. Download the model files from Hugging Face:
   • Visit Hugging Face ESM2 Model.
   • Download the necessary files and place them in the esm_model directory .
5. Download the trained model parameters:
   • If you want to use the fine-tuned model weights directly, visit Zenodo link.
   • Download the trained model parameters and place them in the appropriate directory in your project.

Usage

1. Train Classification Model:

   bash classification.sh

2. Train Regression Model:

    bash regression.sh

3. Run Pipeline: To load the trained models and perform classification and regression to get candiadate seqeunces, run:

    python pipeline.py

   We have already screened the randomly generated sequence space, and the screening results are in the candidate_sequences directory.