Neural Network Approach for Predicting Infrared Spectra from 3D Molecular Structure
Résumé
We developed a machine learning (ML) model to directly predict IR spectra from three-dimensional (3D) molecular structures. The spectra predicted by our model significantly outperform those from density functional theory (DFT) calculations, even after scaling. In a test set of 200 molecules, our model achieves a Spectral Information Similarity Metric (SIS) of 0.92 surpassing the value achieved by DFT scaled frequencies which is 0.57. Additionally, our model considers anharmonic effects offering a fast alternative to laborious anharmonic calculations. Moreover, our model can be used to predict various types of spectra (as UV or NMR) as a function of molecular structure.