This study reports a deep learning approach that utilises message passing neural networks (MPNNs) for predicting chemical shifts in 13C NMR spectra of small molecules. MPNNs were trained on two distinct datasets: one with approximately 4000 labelled structures and another with over 40,000.
In this work, we introduce a novel NMR apodization function designed to enhance spectral resolution while maintaining compatibility with qNMR standards. This function is based on a modified Savitzky–Golay filter, adapted for time-domain application.
The procedure for simulating the nuclear magnetic resonance spectrum linked to the spin system of a molecule for a certain nucleus entails diagonalizing the associated Hamiltonian matrix.