Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/thomasgust/molecumixer
Very incomplete right now, pretrained ARGVAET system for generating, classifying, and predicting the properties of molecules. I couldn't upload the dataset or checkpoints due to size constraints.
https://github.com/thomasgust/molecumixer
argvaet bioinformatics foundation-model generative-ai generative-pretraining gnn molecule neural-network pretraining pytorch rdkit
Last synced: 27 days ago
JSON representation
Very incomplete right now, pretrained ARGVAET system for generating, classifying, and predicting the properties of molecules. I couldn't upload the dataset or checkpoints due to size constraints.
- Host: GitHub
- URL: https://github.com/thomasgust/molecumixer
- Owner: ThomasGust
- License: mit
- Created: 2024-02-02T05:32:36.000Z (9 months ago)
- Default Branch: main
- Last Pushed: 2024-04-24T16:27:35.000Z (6 months ago)
- Last Synced: 2024-10-09T22:05:44.939Z (27 days ago)
- Topics: argvaet, bioinformatics, foundation-model, generative-ai, generative-pretraining, gnn, molecule, neural-network, pretraining, pytorch, rdkit
- Language: Python
- Homepage:
- Size: 13.6 MB
- Stars: 4
- Watchers: 3
- Forks: 1
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Molecumixer
To get this code on GitHub, it was copied from somewhere else on my local machine, so a lot of the paths may be broken. Currently, this project is a work in progress. The main idea is to have a graph neural network encode the molecular strucure and properties of different bioactive molecules found on the internet. Then, using these encodings, we will attempt to decode the original molecule, some molecular characterisitcs and much more. I have ideas for a few novel pretraining tasks that will start to take shape as I write this project out more and more. The main libraries I am using for this project are Numpy, Pytorch Geometric, and RdKit. All of this is written in Python 3. Before this project is finished, if you have any questions or just want to reach out, you can contact me at [email protected].