# yield-rxn

**Repository Path**: dot23/yield-rxn

## Basic Information

- **Project Name**: yield-rxn
- **Description**: Code for the paper: Graph Neural Networks for Predicting Chemical Reaction Performance
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-12-15
- **Last Updated**: 2021-12-15

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README


## Predicting Yield of Chemical Reactions using Deep Learning

### Installation

1. First install Anaconda. 
2. Create a conda environment with
```
conda create --name rxntorch python=3.6
```
3. Then, activate the new conda environment with
```
conda activate rxntorch
```
4. Install RDKit
```
conda install -c rdkit rdkit 
```
5. Installing PyTorch with a CUDA enabled version
```
conda install pytorch torchvision cudatoolkit=10.1 -c pytorch
```
6. Install scikit-learn with
```
conda install scikit-learn
```

Finally, clone this repository to your local machine.

### Getting Started


1. Run: 
```
 pip install requirements
```
2.To train the model, run:
```
python train_yield.py
```
Arguments:
-c: Path to the dataset
--train_split: train-test split ratio
-o: Model name
--epochs: Number of epochs
--seed: Random seed
--layers: Number of layes
--hidden: Hidden size for all layers
--lr_decay: Learning rate decay
--use_domain: Use chemical features or not
--batch_size: Size of mini-batch
--dropout_rate: Droput rate

If using chemical features (domain features) you need the json file containg the features. Otherwise, you can just use smiles strings.