On the value of using 3D-shape and electrostatic similarities in deep generative methods

Multi-parameter optimization, the heart of drug design, is still an open challenge. Thus, improved methods for automated compounds design with multiple controlled properties are desired. Here, we present a significant extension to our previously described fragment-based reinforcement learning method (DeepFMPO) for the generation of novel molecules with optimal properties. As before, the generative process outputs optimized molecules similar to the input structures, now with the improved feature of replacing parts of these molecules with fragments of similar 3D-shape and electrostatics. By performing comparisons of 3D-fragments, we can simulate 3D properties while overcoming the notoriously difficult step of accurately describing bioactive conformations. The comparison of electrostatic potential and molecular shape is performed using the new ESP-Sim python package, allowing the calculation of state-of-the-art partial charges (e.g., RESP with B3LYP/6-31G**) obtained using the quantum chemistry program Psi4. The new improved method is demonstrated with a scaffold-hopping exercise identifying CDK2 bioisosteres. All code is open-source and freely available.

Irradiation-driven molecular dynamics: a review

This paper reviews Irradiation-Driven Molecular Dynamics (IDMD)—a novel computational methodology for atomistic simulations of the irradiation-driven transformations of complex molecular systems implemented in the MBN Explorer software package. Within the IDMD framework, various quantum processes occurring in irradiated systems are treated as random, fast and local transformations incorporated into the classical MD framework in a stochastic manner with the probabilities elaborated on the basis of quantum mechanics. Major transformations of irradiated molecular systems (such as topological changes, redistribution of atomic partial charges, alteration of interatomic interactions) and possible paths of their further reactive transformations can be simulated by means of MD with reactive force fields, in particular with the reactive CHARMM (rCHARMM) force field implemented in MBN Explorer. This paper reviews the general concept of the IDMD methodology and the rCHARMM force field and provides several exemplary case studies illustrating the utilization of these methods. Graphic abstract

Binding Between Cyclohexanohemicucurbit[n]urils and Polar Organic Guests

Inherently chiral, barrel-shaped, macrocyclic hosts such as cyclohexanohemicucurbit[n]urils (cycHC[n]) bind zinc porphyrins and trifluoroacetic acid externally in halogenated solvents. In the current study, we tested a set of eighteen organic guests with various functional groups and polarity, namely, thiophenols, phenols, and carboxylic and sulfonic acids, to identify a preference toward hydrogen bond–donating molecules for homologous cycHC[6] and cycHC[8]. Guests were characterized by Hirshfeld partial charges on acidic hydrogens and their binding by 1H and 19F NMR titrations. Evaluation of association constants revealed the complexity of the system and indirectly proved an external binding with stoichiometry over 2:1 for both homologs. It was found that overall binding strength is influenced by the stoichiometry of the formed complexes, the partial atomic charge on the hydrogen atom of the hydrogen bond donor, and the bulkiness of the guest. Additionally, a study on the formation of complexes with halogen anions (Cl− and Br−) in methanol and chloroform, analyzed by 1H NMR, did not confirm complexation. The current study widens the scope of potential applications for host molecules by demonstrating the formation of hydrogen-bonded complexes with multisite hydrogen bond acceptors such as cycHC[6] and cycHC[8].

Selective Bromocyclization of 5-Amino-4-Alkenyl-1,2,4-Triazole-3-Thione

Here, we present a study on the regioselectivity cyclization of 5-amino-4-alkenyl-1,2,4-triazole-3-thiones. The presence of various nucleophilic centers causes the possibility of cyclization of an alkenyl fragment on different heteroatoms and the formation of a few alternative structures. Elemental bromine was utilized as an electrophilic agent, and two 6-(bromomethyl)-6-R-5,6-dihydro[1,3]thiazolo[2,3-c][1,2,4]triazol-3-amine hydrobromide salts were obtained as the only products when taking reaction in chloroform, acetic acid, or acetonitrile. The 1H and 13C APT NMR spectra analysis proved the formation of the 1,3-thiazolinium ring upon cyclization reaction. DFT calculations at the ωB97X-D3/6-311G(d,p) level of theory were utilized to analyze molecular electrostatic potential, electron localization function, and Hirshfeld atomic partial charges the intermediate bromonium cation. These theoretical calculations explain the experimentally observed regioselectivity.

Revised Atomic Charges for OPLS Force Field Model of Poly(Ethylene Oxide): Benchmarks and Applications in Polymer Electrolyte

Poly(ethylene oxide) (PEO)-based polymers are common hosts in solid polymer electrolytes (SPEs) for high-power energy devices. Molecular simulations have provided valuable molecular insights into structures and ion transport mechanisms of PEO-based SPEs. The calculation of thermodynamic and kinetic properties rely crucially on the dependability of the molecular force fields describing inter- and intra-molecular interactions with the target system. In this work, we reparametrized atomic partial charges for the widely applied optimized potentials for liquid simulations (OPLS) force field of PEO. The revised OPLS force field, OPLSR, improves the calculations of density, thermal expansion coefficient, and the phase transition of the PEO system. In particular, OPLSR greatly enhances the accuracy of the calculated dielectric constant of PEO, which is critical for simulating polymer electrolytes. The reparameterization method was further applied to SPE system of PEO/LiTFSI with O:Li ratio of 16:1. Based on the reparametrized partial charges, we applied separate charge-scaling factors for PEO and Li salts. The charge-rescaled OPLSR model significantly improves the resulting kinetics of Li+ transport while maintaining the accurate description of coordination structures within PEO-based SPE. The proposed OPLSR force field can benefit the future simulation studies of SPE systems.

Optimized Atomic Partial Charges and Radii Defined by Radical Voronoi Tessellation of Bulk Phase Simulations

We present a novel method for the computation of well-defined optimized atomic partial charges and radii from the total electron density. Our method is based on a two-step radical Voronoi tessellation of the (possibly periodic) system and subsequent integration of the total electron density within each Voronoi cell. First, the total electron density is partitioned into the contributions of each molecule, and subsequently the electron density within each molecule is assigned to the individual atoms using a second set of atomic radii for the radical Voronoi tessellation. The radii are optimized on-the-fly to minimize the fluctuation (variance) of molecular and atomic charges. Therefore, our method is completely free of empirical parameters. As a by-product, two sets of optimized atomic radii are produced in each run, which take into account many specific properties of the system investigated. The application of an on-the-fly interpolation scheme reduces discretization noise in the Voronoi integration. The approach is particularly well suited for the calculation of partial charges in periodic bulk phase systems. We apply the method to five exemplary liquid phase simulations and show how the optimized charges can help to understand the interactions in the systems. Well-known effects such as reduced ion charges below unity in ionic liquid systems are correctly predicted without any tuning, empiricism, or rescaling. We show that the basis set dependence of our method is very small. Only the total electron density is evaluated, and thus, the approach can be combined with any electronic structure method that provides volumetric total electron densities—it is not limited to Hartree–Fock or density functional theory (DFT). We have implemented the method into our open-source software tool TRAVIS.