GPU-accelerated approximate kernel method for quantum machine learning.

We introduce Quantum Machine Learning (QML)-Lightning, a PyTorch package containing graphics processing unit (GPU)-accelerated approximate kernel models, which can yield trained models within seconds. QML-Lightning includes a cost-efficient GPU implementation of FCHL19, which together can provide energy and force predictions with competitive accuracy on a microsecond per atom timescale. Using modern GPU hardware, we report learning curves of energies and forces as well as timings as numerical evidence for select legacy benchmarks from atomistic simulation including QM9, MD-17, and 3BPA.

Machine learning meets chemical physics.

Over recent years, the use of statistical learning techniques applied to chemical problems has gained substantial momentum. This is particularly apparent in the realm of physical chemistry, where the balance between empiricism and physics-based theory has traditionally been rather in favor of the latter. In this guest Editorial for the special topic issue on "Machine Learning Meets Chemical Physics," a brief rationale is provided, followed by an overview of the topics covered. We conclude by making some general remarks.

FCHL revisited: Faster and more accurate quantum machine learning.

We introduce the FCHL19 representation for atomic environments in molecules or condensed-phase systems. Machine learning models based on FCHL19 are able to yield predictions of atomic forces and energies of query compounds with chemical accuracy on the scale of milliseconds. FCHL19 is a revision of our previous work [F. A. Faber et al., J. Chem. Phys. 148, 241717 (2018)] where the representation is discretized and the individual features are rigorously optimized using Monte Carlo optimization. Combined with a Gaussian kernel function that incorporates elemental screening, chemical accuracy is reached for energy learning on the QM7b and QM9 datasets after training for minutes and hours, respectively. The model also shows good performance for non-bonded interactions in the condensed phase for a set of water clusters with a mean absolute error (MAE) binding energy error of less than 0.1 kcal/mol/molecule after training on 3200 samples. For force learning on the MD17 dataset, our optimized model similarly displays state-of-the-art accuracy with a regressor based on Gaussian process regression. When the revised FCHL19 representation is combined with the operator quantum machine learning regressor, forces and energies can be predicted in only a few milliseconds per atom. The model presented herein is fast and lightweight enough for use in general chemistry problems as well as molecular dynamics simulations.

Blind test of density-functional-based methods on intermolecular interaction energies.

In the past decade, a number of approaches have been developed to fix the failure of (semi)local density-functional theory (DFT) in describing intermolecular interactions. The performance of several such approaches with respect to highly accurate benchmarks is compared here on a set of separation-dependent interaction energies for ten dimers. Since the benchmarks were unknown before the DFT-based results were collected, this comparison constitutes a blind test of these methods.

Properties and reactivity of nucleic acids relevant to epigenomics, transcriptomics, and therapeutics.

Developments in epigenomics, toxicology, and therapeutic nucleic acids all rely on a precise understanding of nucleic acid properties and chemical reactivity. In this review we discuss the properties and chemical reactivity of each nucleobase and attempt to provide some general principles for nucleic acid targeting or engineering. For adenine-thymine and guanine-cytosine base pairs, we review recent quantum chemical estimates of their Watson-Crick interaction energy, π-π stacking energies, as well as the nuclear quantum effects on tautomerism. Reactions that target nucleobases have been crucial in the development of new sequencing technologies and we believe further developments in nucleic acid chemistry will be required to deconstruct the enormously complex transcriptome.

Accurate ab initio energy gradients in chemical compound space.

Analytical potential energy derivatives, based on the Hellmann-Feynman theorem, are presented for any pair of isoelectronic compounds. Since energies are not necessarily monotonic functions between compounds, these derivatives can fail to predict the right trends of the effect of alchemical mutation. However, quantitative estimates without additional self-consistency calculations can be made when the Hellmann-Feynman derivative is multiplied with a linearization coefficient that is obtained from a reference pair of compounds. These results suggest that accurate predictions can be made regarding any molecule's energetic properties as long as energies and gradients of three other molecules have been provided. The linearization coefficent can be interpreted as a quantitative measure of chemical similarity. Presented numerical evidence includes predictions of electronic eigenvalues of saturated and aromatic molecular hydrocarbons.