The main protease (Mpro) of SARS-CoV-2 is central to viral maturation and is a promising drug target, but little is known about structural aspects of how it binds to its...

Predictive molecular simulations require fast, accurate and reactive interatomic potentials. Machine learning offers a promising approach to construct such potentials by fitting energies and forces to high-level quantum-mechanical data, but...

The main protease (Mpro) of SARS-CoV-2 is central to its viral lifecycle and is a promising drug target, but little is known concerning structural aspects of how it binds to its 11 natural cleavage sites. We used biophysical and crystallographic data and an array of classical molecular mechanics and quantum mechanical techniques, including automated docking, molecular dynamics (MD) simulations, linear-scaling DFT, QM/MM, and interactive MD in virtual reality, to investigate the molecular features underlying recognition of the natural Mpro substrates. Analyses of the subsite interactions of modelled 11-residue cleavage site peptides, ligands from high-throughput crystallography, and designed covalently binding inhibitors were performed. Modelling studies reveal remarkable conservation of hydrogen bonding patterns of the natural Mpro substrates, particularly on the N-terminal side of the scissile bond. They highlight the critical role of interactions beyond the immediate active site in recognition and catalysis, in particular at the P2/S2 sites. The binding modes of the natural substrates, together with extensive interaction analyses of inhibitor and fragment binding to Mpro, reveal new opportunities for inhibition. Building on our initial Mpro-substrate models, computational mutagenesis scanning was employed to design peptides with improved affinity and which inhibit Mpro competitively. The combined results provide new insight useful for the development of Mpro inhibitors.

Recent advances in the development of reactive machine-learned potentials (MLPs) promise to transform reaction modelling. However, such methods have remained computationally expensive and limited to experts. Here, we employ different...

<p>Predictive simulations of dynamic processes in molecular systems require fast, accurate and reactive interatomic potentials. Machine learning offers a promising approach to construct force-field models for large-scale molecular simulation by fitting to high-level quantum-mechanical data. However, machine-learned force fields generally require considerable human intervention and data volume. Here we show that, by leveraging hierarchical and active learning, accurate Gaussian Approximation Potential (GAP) models for diverse chemical systems can be developed in an autonomous way, requiring only hundreds to a few thousand energy and gradient evaluations on the reference potential-energy surface. Our approach relies on a decomposition of the condensed-phase molecular system into intra- and inter-molecular terms, and on the definition of a prospective error metric to quantify accuracy. We demonstrate applications to a range of molecular systems: from bulk water, organic solvents, and a solvated ion onwards to the description of chemical reactivity, including, a bifurcating Diels–Alder reaction in the gas phase and non-equilibrium dynamics (S_N2 reaction) in explicit solvent. The method provides a route to routinely generating machine-learned force fields for complex and/or reactive molecular systems. </p>