https://github.com/dmis-lab/arpnet

ARPNet: Antidepressant Response Prediction Network for Major Depressive Disorder
https://github.com/dmis-lab/arpnet

Last synced: about 1 month ago
JSON representation

ARPNet: Antidepressant Response Prediction Network for Major Depressive Disorder

• Host: GitHub
• URL: https://github.com/dmis-lab/arpnet
• Owner: dmis-lab
• Created: 2019-01-29T10:36:30.000Z (almost 6 years ago)
• Default Branch: master
• Last Pushed: 2019-01-30T14:10:19.000Z (almost 6 years ago)
• Last Synced: 2024-05-14T00:23:17.034Z (7 months ago)
• Language: Python
• Homepage:
• Size: 170 KB
• Stars: 2
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# ARPNet

![model image](model_description.PNG)

This repository provides implementation code of ARPNet, a Antidepressant Response Prediction Network for Major Depressive Disorder.

## Installation

The experiments were conducted on a single TITAN Xp GPU machine which has 12GB of RAM.

The implemntation code of ARPNet was tested with the following requirements:

*   **[`Python2.7.12`](https://www.python.org/downloads/release/python-2712/)**

*   **[`TensorFlow1.6.0`](https://www.tensorflow.org/install/)**

## Datasets

### Description

We conducted experients on the data of 121 patients with MDD collected from Korea University Anam Hospital, Seoul, Korea. By several feature selection steps, we extracted some useful features, which are demographic, neuroimaging biomarkers, genetic variants, and DNA methylation features, for predicting antidepressant response from patient data. Because of the Korea University Anam Hospital's policies related to the patients' personal information, the data used in the experiment could not be released. 

### Toy Smaples

To run our implementation code, we release the toy samples `\toy\train_1.tsv`, `\toy\train_2.tsv`, `\toy\test_1.tsv`, `\toy\test_2.tsv` , and `\toy\patient_data.tsv`

* `train_1.tsv` and `test_1.tsv` are used for the prediction of the degree of antdepressant response (Task 1). 

* `train_2.tsv` and `test_2.tsv` are used for the prediction of the clinical remession of patients (Task 2).

* `patient_data.tsv` consists of demographic, neuroimaging biomarkers, genetic variants, and DNA methylation features.

## Run

Following command runs ARPNet code on toy samples with default parameters.

```

python run.py \

    --task=1

```

You can change the task as you want. Once you perfrom task 2, you can use it in task mode by using `--task=2`

## Citation

If we submit the paper to a conference or journal, we will update the BibTeX.

## Contact information

For help or issues using ARPNet, please submit a GitHub issue. Please contact Buru Chang

(`buru_chag (at) korea.ac.kr`), or Yonghwa Choi (`yonghwachoi (at) korea.ac.kr`) for communication related to ARPNet.