maozhitao/bere

BERE

Implementation of the paper [BERE: A novel machine learning framework for accurate biomedical entity relation extraction].

Environments

Tested on a linux server with GeForce GTX 1080 and the running environment is as follows:

• Python 3.5.2

• PyTorch 1.0.0

• sklearn 0.20.2

• numpy 1.15.4

• cuda 9.0

Installation Guide

1. Download the pretrained word embedding PubMed-and-PMC-w2v.bin from http://evexdb.org/pmresources/vec-space-models/ and put it in ./data/.

2. Download the complete DTI dataset from https://cloud.tsinghua.edu.cn/d/1bdb3bed3031479c8aa9/ and put it in ./data/dti/.

How to Run

[DDI Expirement](less than 1h for one training)

1. Run ./data/ddi/data_prepare.py to preprocess the DDI dataset.

2. Run ./train_ddi.py to train BERE with different learning rates.

3. Run ./test_ddi.py to test BERE with the best model.

 

[DTI Expirement](taking 10~20h before convergence)

1. Run ./data/dti/data_prepare.py to preprocess the DTI dataset.

2. Run ./train_dti.py to train BERE with different learning rates.

3. Run ./test_dti.py to test BERE with the best model.

 

[Demo of DTI Prediction]

1. Train the model on the DTI dataset.

2. Run ./data/dti/transform.py to preprocess the pmc_nintedanib dataset.

3. Run ./predict.py to predict the targets of the drug nintedanib by the well-trained model.

Data Description

• PubMed-and-PMC-w2v.bin: The pretrained word embedding.
• train.json, valid.json, test.json: The original dataset.
• label2id.json: The label file.
• pmc_nintedanib.json: The data for DTI Prediction demo.
• tree_examples.json: The data for visualization demo.
• config.py: The hyper-parameter settings.

File Description

• ./data/: This directory contains DDI dataset, DTI dataset and pretrained word embedding.

• ./network/: This directory contains the codes of our model.

• ./checkpoint/(generated): This directory contains the checkpoints of model in the training process.

• ./result/(generated): This directory contains the test results and prediction results

• ./output/(generated): This directory contains the prediction results, which is more convenient for reading.

• ./train_ddi.py: This is a demo for training the BERE on DDI dataset.

• ./train_dti.py: This is a demo for testing the BERE on DDI dataset.

• ./test_ddi.py: This is a demo for training the BERE on DTI dataset.

• ./test_dti.py: This is a demo for testing the BERE on DTI dataset.

• ./predict.py: This is a demo for predicting the targets of the drug nintedanib.

• ./plot_pr.py: This file is used to plot the precision-recall curve of the results in ./result/.

• ./visualize.py(optional): This file is used for the visualization of word attention, sentence attention and sentence tree structures.

Notes

• The full backup data, codes and results can be found in https://cloud.tsinghua.edu.cn/d/0e3a253403914c33b3dd/, which could be used for reproduction.
• The full datasets for discovering novel DTIs is available from the corresponding authors upon request.
• If you have any other questions or comments, please feel free to email Lixiang Hong (honglx17[at]mails[dot]tsinghua[dot]edu[dot]cn) and/or Jianyang Zeng (zengjy321[at]tsinghua[dot]edu[dot]cn).