SPICE Dataset Released
A new DFT dataset for small molecules and more
The SPICE dataset is a collection of over 1.1 million molecular structures used to train machine learning models for simulating drug-like molecules and peptides interacting with proteins. The dataset includes a diverse range of molecules and interactions and provides forces and energies calculated at a high level of theory. By using the SPICE dataset to train machine learning potentials, researchers can improve the accuracy and efficiency of molecular simulations in drug discovery and other applications. Click here to know more…