BERT-TCR

https://github.com/zhangbeibei-min/berttcr

Admin

A deep learning framework for cancer-related immune status prediction using TCR repertoire data

Tokens:365

Snippets:5

Trust Score:4

Update:1 year ago

Show doc for...

Tokens:

Raw

### Install BertTCR Dependencies

Source: https://github.com/zhangbeibei-min/berttcr/blob/main/README.md

Command to install all necessary Python packages required to run BertTCR. This command reads the requirements from the 'requirements.txt' file.

```bash
pip install -r requirements.txt
```

--------------------------------

### Clone BertTCR Repository

Source: https://github.com/zhangbeibei-min/berttcr/blob/main/README.md

Instructions to clone the BertTCR GitHub repository and navigate into the project directory. This is the first step for setting up the project locally.

```bash
git clone https://github.com/zhangbeibei-min/BertTCR.git
cd BertTCR
```

--------------------------------

### Predict Immune Status using Pre-trained Model

Source: https://github.com/zhangbeibei-min/berttcr/blob/main/README.md

Run prediction using a pre-trained BertTCR model. This script takes the path to a pre-trained model file as input and outputs prediction results to a TSV file.

```python
python .\Codes\BertTCR_prediction.py --model_file ./TrainedModels/Pretrained_THCA.pth
```

--------------------------------

### Train a New BertTCR Model

Source: https://github.com/zhangbeibei-min/berttcr/blob/main/README.md

Command to train a new BertTCR model using custom TCR sequencing data. Users can specify training and validation data directories, dropout rate, number of epochs, and learning rate.

```python
python BertTCR_tranining.py --sample_dir ./TrainingData --val_sample_dir ./validationData --dropout 0.4 --epoch 500 --learning_rate 0.001
```

--------------------------------

### Evaluate BertTCR Prediction Performance

Source: https://github.com/zhangbeibei-min/berttcr/blob/main/README.md

Evaluate the performance of BertTCR predictions using a specific output file. This script calculates metrics such as Accuracy, Sensitivity, Specificity, F1-score, and AUC.

```python
./Codes/BertTCR_evalution.py
```

Tokens:

Raw