File size: 1,061 Bytes
7a096a1
24eaaa7
 
 
 
 
 
 
7a096a1
 
 
 
 
95f97c5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
---
title: ReactXT
emoji: 🧪
colorFrom: yellow
colorTo: green
sdk: gradio
sdk_version: 4.36.0
python_version: 3.8
app_file: app.py
pinned: false
---


# ReactXT: Understanding Molecular “Reaction-ship” via Reaction-Contextualized Molecule-Text Pretraining

## Comparison to previous molecule-text generative modeling methods

![fig1](./figures/comparison.pdf)


## Framework of ReactXT

![fig1](./figures/frameworks.pdf)


## Requirements

Our environment is detailed in `environment.yml`. To create a new environment `reactxt`, run the following command:

```bash
conda env create -f environment.yml
```


## Reproduce the results

### Reaction-Contextualized Molecule-Text Pretraining

```bash
bash scripts/run_pretrain.sh
```

### Finetuning on downstream tasks

1. Experimental Procedure Prediction on OpenExp

```bash
bash scripts/run_action.sh
```

2. Molecule Captioning on PubChem324k and CheBI-20

```bash
bash scripts/run_caption.sh
bash scripts/run_chebi.sh
```

3. Retro-synthesis Prediction on USPTO-50k

```bash
bash scripts/run_retro.sh
```