https://github.com/ds4sd/molannotator
https://github.com/ds4sd/molannotator
Last synced: 12 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/ds4sd/molannotator
- Owner: DS4SD
- License: mit
- Created: 2024-02-16T16:02:36.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2024-07-08T16:00:44.000Z (about 2 years ago)
- Last Synced: 2025-03-16T13:35:40.604Z (over 1 year ago)
- Language: Python
- Size: 48.8 KB
- Stars: 7
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
- Code of conduct: CODE_OF_CONDUCT.md
Awesome Lists containing this project
README
# MolAnnotator
This repository implements applications to ease manual annotation of chemical images.
## Installation
```
conda create -n molannotator python=3.11
conda activate molannotator
pip install -e .
```
## Classification
The classification application allows to assign to an image one of the following categories: `standard_molecule` (Molecular Structure), `markush_structure` (Markush Structure), `miscellaneous` (Background), `polymer` (Markush Structure), and `segmentation_annotation_error`.
To run the application:
1. Move the images to be annotated to `./data/classfication/default/images/*.png`,
2. (optional) Move pre-annotations to `./data/classfication/default/classes/*.json`,
3. Run `bash ./molannotator/apps/run_classification_app.sh`.
Annotations are saved in `./data/classfication/default/classes/*.json`.
## Recognition
The recognition application allows to annotate the molecular graph of a chemical-structure image.
To run the application:
1. Move the images to be annotated to `./data/recognition/default/images/*.png`,
2. (optional) Create pre-annotations by running `python3 ./molannotator/molscribe_pre_annotate.py` (see [MolScribe](https://github.com/thomas0809/MolScribe) for installation steps),
3. (optional) Move pre-annotations to `./data/recognition/default/predictions/*.mol`,
3. Run `bash ./molannotator/apps/run_recognition_app.sh`.
Annotations are saved in `./data/recognition/default/molfiles/*.mol`.