Quantum chemical benchmark databases of gold-standard dimer interaction energies

Donchev, Alexander; Taube, Andrew G.; Decolvenaere, Elizabeth; Hargus, Cory; McGibbon, Robert T.; Law, Ka-Hei; Gregersen, Brent A.; Li, Je-Luen; Palmö, Kim; Siva, Karthik; Bergdorf, Michael; Klepeis, John L.; Shaw, David E.

doi:10.1038/s41597-021-00833-x

Cited by 76 publications

(81 citation statements)

References 94 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Treating noncovalent interaction (NCI) within QML is an interesting and important research subject, with relevant large data sets emerging only as of recently. Most notably, several collections of NCI data sets have by now become publicly available, 381 covering 3700 distinct types of interacting molecule pairs: (i) DES370 K, contains interaction energies for more than 370k dimer geometries with NCI energy calculated at the level of CCSD(T)/CBS (MP2(aVTZ, aVQZ) correlation energy is used for extrapolation, and (ii) DES5M, comprising NCI energies calculated using SNS-MP2, for nearly 5 M dimer geometries. The monomers involved include typical organic species, made up of common p-block elements as well as alkali metal ions, most of which containing no more than seven heavy atoms.…”

Section: Data Setsmentioning

confidence: 99%

Ab Initio Machine Learning in Chemical Compound Space

2021

View full text Add to dashboard Cite

Chemical compound space (CCS), the set of all theoretically conceivable combinations of chemical elements and (meta-)stable geometries that make up matter, is colossal. The first-principles based virtual sampling of this space, for example, in search of novel molecules or materials which exhibit desirable properties, is therefore prohibitive for all but the smallest subsets and simplest properties. We review studies aimed at tackling this challenge using modern machine learning techniques based on (i) synthetic data, typically generated using quantum mechanics based methods, and (ii) model architectures inspired by quantum mechanics. Such Quantum mechanics based Machine Learning (QML) approaches combine the numerical efficiency of statistical surrogate models with an ab initio view on matter. They rigorously reflect the underlying physics in order to reach universality and transferability across CCS. While state-of-the-art approximations to quantum problems impose severe computational bottlenecks, recent QML based developments indicate the possibility of substantial acceleration without sacrificing the predictive power of quantum mechanics.

show abstract

Section: Data Setsmentioning

confidence: 99%

Ab Initio Machine Learning in Chemical Compound Space

2021

View full text Add to dashboard Cite

show abstract

“…h) defined as the difference between the energy of a molecule deformed along a particular normal mode and the energy of the same molecule at equilibrium. HW6Cl-anionic 215,216 , HW6Fanionic 215,216 , FmH2O10anionic 215,216 , SW49Bind345anionic 217 , SW49Bind6anionic 217 , Anionpi-anionic 218 , IL236-anionic 219 , DES15Kanionic 206 , NENCI-2021anionic 207 , CHAL336anionic 212 , XB45-anionic 211 +0.17 to +33.79 negatively charged molecules a) defined as the difference between the energy of the complex and the sum of energy of the monomers. A negative interaction energy indicates the complex is more stable than the separated monomers.…”

Section: Training and Validation Data Setsmentioning

confidence: 99%

“…We now explore the same idea by taking some subsets of the validation set and purposefully applying only part of the available ACPs so that not all atoms in the system have an associated correction. Table 6 presents a summary of the MAEs using various methods for two data sets from the validation set: the DES15K 206 set of non-covalent interaction energies (11474 data points) and the Torsion30 228 set of molecular conformational energies (2107 data points). The table shows that the MAEs of the HF/MINIX method are 4.83 and 0.92 kcal/mol for the DES15K and Torsion30 data sets, respectively.…”

Section: Applications Of Acps Developed For Hf-3cmentioning

confidence: 99%

Fast and accurate quantum mechanical modeling of large molecular systems using small basis set Hartree–Fock methods corrected with atom-centered potentials

Prasad

Otero-de-la-Roza

DiLabio

2022

Preprint

View full text Add to dashboard Cite

There has been significant interest in developing fast and accurate quantum mechanical methods for modeling large molecular systems. In this work, by utilizing a machine-learning regression technique, we have developed new low-cost quantum mechanical approaches to model large molecular systems. The developed approaches rely on using one-electron Gaussian-type functions called atom-centered potentials (ACPs) to correct for the basis set incompleteness and the lack of correlation effects in the underlying minimal or small basis set Hartree-Fock (HF) methods. In particular, ACPs are proposed for ten elements common in organic and bio-organic chemistry (H, B, C, N, O, F, Si, P, S, and Cl) and four different base methods: two minimal basis sets (MINIs and MINIX) plus a double-ζ basis set (6-31G*) in combination with dispersion-corrected HF (HF-D3/MINIs, HF-D3/MINIX, HF-D3/6-31G*), and the HF-3c method. The new ACPs are trained on a very large set (73832 data points) of non-covalent properties (interaction and conformational energies) and validated additionally on a set of 32048 data points. All reference data is of complete basis set coupled-cluster quality, mostly CCSD(T)/CBS. The proposed ACP-corrected methods are shown to give errors in the tenths of a kcal/mol range for non-covalent interaction energies and up to 2 kcal/mol for molecular conformational energies. More importantly, the average errors are similar in the training and validation sets, confirming the robustness and applicability of these methods outside the boundaries of the training set. In addition, the performance of the new ACP-corrected methods is similar to complete basis set DFT but at a cost that is orders of magnitude lower, and the proposed ACPs can be used in any computational chemistry program that supports effective-core potentials without modification. It is also shown that ACPs improve the description of covalent and non-covalent bond geometries of the underlying methods and that the improvement brought about by the application of the ACPs is directly related to the number of atoms to which they are applied, allowing the treatment of systems containing some atoms for which ACPs are not available. Overall, the ACP-corrected methods proposed in this work constitute an alternative accurate, economical, and reliable quantum mechanical approach to describe the geometries, interaction energies, and conformational energies of systems with hundreds to thousands of atoms.

show abstract

“…This implies that benchmarking is often a necessary pre-step for HTVS, whether it is done by the researchers carrying out the screening or provided by other benchmarking studies in the literature. There have been multiple recent works [75][76][77][78][79] with the purpose of accelerating the future computation of a specific property by comparison, modification, and combination of existing theory.…”

Section: Computable Properties: Benchmarking and Calibrationmentioning

confidence: 99%