This is the repository of the framework related to the work "Fast Evaluation of the Adsorption Energy of Organic Molecules on Metals via Graph Neural Networks", preprint here. The graph neural networks developed within this framework allow the fast prediction of the DFT ground state energy of the following systems: All closed-shell molecules containing C, H, O, N and S. Same molecules mentioned above adsorbed on the following 12 transition metals: Ag, Au, Cd, Cu, Ir, Ni, Os, Pd, Pt, Rh, Ru, Zn. The framework is built with PyTorch and PyTorch Geometric.

Includes a minimal version of the FG- and BM-datasets.