can read csv (position)file？

XieResearchGroup / Physics-aware-Multiplex-GNN

Code for our Nature Scientific Reports paper "A universal framework for accurate and efficient geometric deep learning of molecular systems"

https://www.nature.com/articles/s41598-023-46382-8

53 stars 11 forks source link

can read csv (position)file？ #2

Closed BonedDeng closed 8 months ago

zetayue commented 8 months ago

Yes, as long as your csv file contains the necessary information to be loaded as the data object in PyTorch Geometric. The current dataloaders in this repo are built to load the data format in the benchmark datasets.

BonedDeng commented 8 months ago

but it need smiles and 3dcoordinate as input？

zetayue commented 8 months ago

SMILES is not necessary since it is mainly for small molecules. Any data format (mol, pdb, sdf, etc.) that can provide 3d coordinates and chemical features to represent a 3D molecular structure is okay. The dataloader has to be modified to load your own data as the data object in PyG.