xwang38438 / Cell-Attention-Network-on-Molecule

Test if higher-order data structure can perform better than general GNNs on molecular graphs

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Cell-Attention-Network-on-Molecule

Test if higher-order data structure can perform better than general GNNs on molecular graphs

Installation

  1. follow the instruction in TopoModelX
  2. install graph-tool by conda install graph-tool=2.58=py311h35b2f40_1 -c conda-forge

Data Preprocessing

To preprocess the molecular graph datasets for training models, run the following command:

python data/preprocess.py --dataset QM9 python data/preprocess_for_nspdk.py --dataset QM9

and then run the cells in ``

About

Test if higher-order data structure can perform better than general GNNs on molecular graphs


Languages

Language:Jupyter Notebook 78.1%Language:Python 21.9%