Semiempirical quantum mechanical methods with corrections for noncovalent interactions, namely dispersion and hydrogen bonds, reach an accuracy comparable to much more expensive methods while being applicable to very large systems (up to 10 000 atoms). These corrections have been successfully applied in computer-assisted drug design, where they significantly improve the correlation with the experimental data. Despite these successes, there are still several unresolved issues that limit the applicability of these methods. We introduce a new generation of both hydrogen-bonding and dispersion corrections that address these problems, make the method more robust, and improve its accuracy. The hydrogen-bonding correction has been completely redesigned and for the first time can be used for geometry optimization and molecular-dynamics simulations without any limitations, as it and its derivatives have a smooth potential energy surface. The form of this correction is simpler than its predecessors, while the accuracy has been improved. For the dispersion correction, we adopt the latest developments in DFT-D, using the D3 formalism by Grimme. The new corrections have been parametrized on a large set of benchmark data including nonequilibrium geometries, the S66x8 data set. As a result, the newly developed D3H4 correction can accurately describe a wider range of interactions. We have parametrized this correction for the PM6, RM1, OM3, PM3, AM1, and SCC-DFTB methods.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1021/ct200751e | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Enhancing Activation Energy Predictions under Data Constraints Using Graph Neural Networks.
J Chem Inf Model
January 2025
Department of Chemical Engineering, National Taiwan University, No. 1, Section 4, Roosevelt Road, Taipei 10617, Taiwan.
Accurately predicting activation energies is crucial for understanding chemical reactions and modeling complex reaction systems. However, the high computational cost of quantum chemistry methods often limits the feasibility of large-scale studies, leading to a scarcity of high-quality activation energy data. In this work, we explore and compare three innovative approaches (transfer learning, delta learning, and feature engineering) to enhance the accuracy of activation energy predictions using graph neural networks, specifically focusing on methods that incorporate low-cost, low-level computational data.
View Article and Find Full Text PDFFullerene-Doped Poly(ionic liquids) as Small Molecular Gas Sensors-Control of Intermolecular Interactions.
ACS Omega
January 2025
Department of Functional Materials, FZU - Institute of Physics - Czech Academy of Sciences, Na Slovance 1999/2, Prague 8 182 00, Czech Republic.
Here, we investigate the interactions between five representative gaseous analytes and two poly(ionic liquids) (PILs) based on the sulfopropyl acrylate polyanion in combination with the alkylphosphonium cations, P and P, and their nanocomposites with fullerenes (C, C) to reveal the potential of PILs as sensitive layers for gas sensors. The gaseous analytes were chosen based on their molecular size (all of them containing two carbon atoms) and variation of functional groups: alcohol (ethanol), nitrile (acetonitrile), aldehyde (acetaldehyde), halogenated alkane (bromoethane), and carboxylic acid (acetic acid). The six variations of PILs-PSPA (), PSPA + C ( + C), PSPA + C ( + C), and PSPA (), PSPA + C ( + C), PSPA + C ( + C)-were characterized by UV-vis and Raman spectroscopy, and their interactions with each gaseous analyte were studied using electrochemical impedance spectroscopy.
View Article and Find Full Text PDFCreation of an FeP Schreibersite Density Functional Tight Binding Model for Astrobiological Simulations.
J Phys Chem A
January 2025
Physical and Life Sciences Directorate, Lawrence Livermore National Laboratory, 7000 East Avenue, Livermore, California 94550, United States.
The mineral schreibersite, e.g., FeP, is commonly found in iron-rich meteorites and could have served as an abiotic phosphorus source for prebiotic chemistry.
View Article and Find Full Text PDFEnd-Point Affinity Estimation of Galectin Ligands by Classical and Semiempirical Quantum Mechanical Potentials.
J Chem Inf Model
January 2025
Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences, Gilead Sciences & IOCB Research Centre, Flemingovo nám. 2, 166 10 Prague, Czech Republic.
The use of quantum mechanical potentials in protein-ligand affinity prediction is becoming increasingly feasible with growing computational power. To move forward, validation of such potentials on real-world challenges is necessary. To this end, we have collated an extensive set of over a thousand galectin inhibitors with known affinities and docked them into galectin-3.
View Article and Find Full Text PDFPM6-ML: The Synergy of Semiempirical Quantum Chemistry and Machine Learning Transformed into a Practical Computational Method.
J Chem Theory Comput
January 2025
Institute of Organic Chemistry and Biochemistry, Czech Academy of Sciences, 160 00 Prague, Czech Republic.
Machine learning (ML) methods offer a promising route to the construction of universal molecular potentials with high accuracy and low computational cost. It is becoming evident that integrating physical principles into these models, or utilizing them in a Δ-ML scheme, significantly enhances their robustness and transferability. This paper introduces PM6-ML, a Δ-ML method that synergizes the semiempirical quantum-mechanical (SQM) method PM6 with a state-of-the-art ML potential applied as a universal correction.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!