Many molecular design tasks benefit from fast and accurate calculations of quantum-mechanical (QM) properties. However, the computational cost of QM methods applied to drug-like molecules currently renders large-scale applications of quantum chemistry challenging. Aiming to mitigate this problem, we developed DelFTa, an open-source toolbox for the prediction of electronic properties of drug-like molecules at the density functional (DFT) level of theory, using Δ-machine-learning. Δ-Learning corrects the prediction error (Δ) of a fast but inaccurate property calculation. DelFTa employs state-of-the-art three-dimensional message-passing neural networks trained on a large dataset of QM properties. It provides access to a wide array of quantum observables on the molecular, atomic and bond levels by predicting approximations to DFT values from a low-cost semiempirical baseline. Δ-Learning outperformed its direct-learning counterpart for most of the considered QM endpoints. The results suggest that predictions for non-covalent intra- and intermolecular interactions can be extrapolated to larger biomolecular systems. The software is fully open-sourced and features documented command-line and Python APIs.
1. The Δ-machine learning models within DelFTa have been extended for the predictions of (i) non-covalent intra- and intermolecular interactions, and (ii) relative formation energies for different conformers of the same molecule. 2. Additional benchmark experiments were conducted, including predictions for (i) non-covalent bonds in biomolecular systems (such as DNA and RNA base pairs, ligand-protein binding pockets, alpha-helices and beta-sheets), and (ii) quantum properties of charged molecules. 3. Pooling operations for extensive and intensive quantum properties have been adapted to enable optimal prediction accuracies for molecular properties (absolute and relative formation energies, orbital energies and dipoles).