Flexibilities of wavelets as a computational basis set for large-scale electronic structure calculations

Abstract

The BigDFT project started in 2005 with the aim of testing the advantages of using a Daubechies wavelet basis set for Kohn-Sham density functional theory with pseudopotentials. This project led to the creation of the BigDFT code, which employs a computational approach with optimal features for exibility, performance and precision of the results. In particular, the employed formalism has enabled the implementation of an algorithm able to tackle DFT calculations of large systems, up to many thousands of atoms, with a computational eort which scales linearly with the number of atoms. In this work we recall some of the features that have been made possible by the peculiar properties of Daubechies wavelets. In particular, we focus our attention on the usage of DFT for large-scale systems. We show how the localised description of the KS problem, emerging from the features of the basis set, are helpful in providing a simplied description of large-scale electronic structure calculations. We provide some examples on how such simplied description can be employed, and we consider, among the case-studies, the SARS-CoV-2 main protease.

Mechanistic Insight from Full Quantum Mechanical Modeling: Laccase as a Detoxifier of Aflatoxins

“…To test if this assumption is valid, from the experimental data we define detoxification efficiency ≝ (2), and using the value of estimated from experimental data, the model accurately approximates the measured kinetics in the case of AFG2 throughout the experimental time ( Fig. 1D), further confirming that this model is suitable for representing aflatoxin detoxification by laccase.…”

“…In our earlier publications [2], [23], we described a Complexity Reduction framework which uses the electronic density computed by QM calculations to reproduce the properties of a full system from calculations on only a subset of the system. The key ingredient of this analysis is the Fragment Bond Order, which is a generalization of atomic bond order to interactions between two arbitrary sets of atoms.…”

“…Our model, encompassing all laccase's amino acid residues plus the aflatoxin molecule, quantum mechanically simulates approximately 7600 atoms. This methodological development is implemented using the BigDFT code [2].…”

Microscopic Factors Modulating the Interactions Between the SARS-CoV-2 Main Protease and α−Ketoamide Inhibitors

“…are the most reliable theoretical approaches to investigate microscopic systems. Recently one of us proposed an efficient quantum O(N ) method based on Daubechies wavelets 6,7 allowing one to investigate molecular systems comprising thousands of atoms on modern supercomputing systems. An interesting feature of that approach is its ability to decompose a molecular system into fragments from which one may compute a map summarizing the microscopic local interactions occurring within a molecular complex.…”

“…Along the last 2 ns MD segments, we extracted 100 M pro /inhibitor regularly spaced snapshots (each comprising about 4 800 atoms) that were further investigated using a full QM O(N ) approach based on Density Functional Theory with the Perdew-Burke-Erzerhof (PBE) functional implemented by a Daubechies wavelets formalism. 6,7 On modern supercomputing systems the computation of a M pro /inhibitor complex PBE single energy point is achieved within less than 2 hours using 1 024 computational cores. Note that the PBE energies discussed below (unless otherwise stated) have been corrected by including D3 dispersion terms.…”