Official implementation of pre-training via denoising for TorchMD-NET

🥇Samsung AI Challenge 2021 1등 솔루션입니다🥇

[KDD'22] Source codes of "Graph Rationalization with Environment-based Augmentations"

An atom-bond transformer-based message passing neural network for molecular property prediction.

[ICML 2023] Hierarchical Grammar-Induced Geometry for Data-Efficient Molecular Property Prediction

Predict optical properties of molecules with machine learning.

The official open-source repository for AutoMolDesigner, an easy-to-use Python application dedicated to automated molecular design.

An efficient curriculum learning-based strategy for molecular graph learning

Code and Data for the paper: Graph Sampling-based Meta-Learning for Molecular Property Prediction [IJCAI2023]

Exploring QSAR Models for Activity-Cliff Prediction

The code base for AWARE, a graph representation learning method published at TMLR

MUBen: Benchmarking the Uncertainty of Molecular Representation Models

3rd place solution for 2022 Samsung AI Challenge (Materials Discovery)

IUPAC-based large-scale molecular pre-trained model for property prediction and molecular generation

KDD-23 Automated 3D Pre-Training for Molecular Property Prediction

This repository is a brief tutorial about how Graph convolutional networks and message passing networks work with example code demonstration using pytorch and torch_geometric

Graduation Design

Samsung AI Challenge for Scientific Discovery, Samsung Advanced Institute of Technology and Dacon, ~2021.09.27

Package for TwinBooster. Enables fast and powerful zero-shot molecular property prediction.

Molecular-property prediction with sparsity