Linear-scaling density-functional theory with tens of thousands of atoms: Expanding the scope and scale of calculations with ONETEP

Hine, Nicholas D. M.; Haynes, Peter; Mostofi, Arash A.; Skylaris, Chris‐Kriton; Payne, Mike C.

doi:10.1016/j.cpc.2008.12.023

Cited by 130 publications

(142 citation statements)

References 46 publications

Supporting

Mentioning

142

Contrasting

Order By: Relevance

“…In particular, Kannappan and Gready 15 have made impressive progress in this field with their seminal work on molecular modelling of various reactive pathways in Rubisco. However, the increase in computer power in the past few decades, and, crucially, algorithmic advances in linear-scaling DFT 17 , permits the accurate study of larger system sizes to thousands of atoms 18,19 , and the modelling of entire proteins 20,21 . This allows for the inclusion of the protein matrix well beyond the active site, which QM and QM/MM neglect by necessity, due to its computationally prohibitive nature.…”

Section: Introductionmentioning

confidence: 99%

Comparative studies for evaluation of CO2 fixation in the cavity of the Rubisco enzyme using QM, QM/MM and linear-scaling DFT methods

2013

View full text Add to dashboard Cite

We evaluate the minimum energy configuration (MM) and binding free energy (QM/MM and QM) of CO 2 to Rubisco, of fundamental importance to the carboxylation step of the reaction catalysed by Rubisco. Two structural motifs have been used to achieve this goal, one of which starts from the initial X-ray Protein Data Bank structure of Rubisco's active centre (671 atoms), and the other is a simplified, smaller model (77 atoms) which has been used most successfully, thus far, for study. The small model is subjected to quantum chemical Density functional theory (DFT) studies, both in vacuo and using implicit solvation. The effects of the protein environment are also included by means of a hybrid quantum mechanical/molecular mechanical (QM/MM) approach, using PM6/AMBER and B3LYP/AMBER schemes. Finally, linear-scaling DFT methods have also been applied to evaluate energetic features of the large motif, and the result obtained for the binding free energy of the CO 2 underlines the importance of the accurate modelling of the surrounding protein milieu using a full DFT description.

show abstract

Section: Introductionmentioning

confidence: 99%

Comparative studies for evaluation of CO2 fixation in the cavity of the Rubisco enzyme using QM, QM/MM and linear-scaling DFT methods

2013

View full text Add to dashboard Cite

show abstract

“…ONETEP therefore combines the benefits of linear-scaling with system size with the variational bounds and systematic convergence with respect to basis size provided by a planewave basis. Recent development work 13 on the ONETEP code resulted in considerable speedup to the performance of LS-DFT simulations. However, it was also demonstrated that sparse algebra operations remained the limiting factor on parallel efficiency when scaling to large numbers of processes and that scaling of the calculation wall-clock time as O͑N / P͒ was not obtained beyond around P ϳ 100.…”

Section: A Sparse Matrices In Electronic Structurementioning

confidence: 99%

“…13 It uses a set of optimizable localized functions, referred to as nonorthogonal generalized Wannier functions ͑NGWFs͒, 14 expressed in terms of periodic sinc functions ͑psincs͒, 15 to represent the density matrix. The basis of psinc functions has very similar favorable properties to the plane-wave basis frequently used in traditional DFT.…”

Section: A Sparse Matrices In Electronic Structurementioning

confidence: 99%

Linear-scaling density-functional simulations of charged point defects in Al2O3 using hierarchical sparse matrix algebra

Hine

Haynes

Mostofi

et al. 2010

The Journal of Chemical Physics

View full text Add to dashboard Cite

We present calculations of formation energies of defects in an ionic solid ͑Al 2 O 3 ͒ extrapolated to the dilute limit, corresponding to a simulation cell of infinite size. The large-scale calculations required for this extrapolation are enabled by developments in the approach to parallel sparse matrix algebra operations, which are central to linear-scaling density-functional theory calculations. The computational cost of manipulating sparse matrices, whose sizes are determined by the large number of basis functions present, is greatly improved with this new approach. We present details of the sparse algebra scheme implemented in the ONETEP code using hierarchical sparsity patterns, and demonstrate its use in calculations on a wide range of systems, involving thousands of atoms on hundreds to thousands of parallel processes.

show abstract

“…The atoms are then ordered in such a way that the dense matrix blocks of P {v} are clustered around the diagonal. Both the exact knowledge of the sparsity pattern and the ordering of the atoms before the start of a calculation allows for optimised communication and efficient sharing matrix data over the CPUs available to the calculation [106].…”

Section: Sparsity and Linear Scalingmentioning

confidence: 99%

“…Together with the localisation radius r c on the NGWFs, the cutoff radius on the density matrix elements predefines the sparsity pattern of any matrix encountered in a ground state calculation [106]. The atoms are then ordered in such a way that the dense matrix blocks of P {v} are clustered around the diagonal.…”

Section: Sparsity and Linear Scalingmentioning

confidence: 99%

Computing the Optical Properties of Large Systems

Zuehlsdorff

2015

Springer Theses

View full text Add to dashboard Cite

Linear-scaling density-functional theory with tens of thousands of atoms: Expanding the scope and scale of calculations with ONETEP

Cited by 130 publications

References 46 publications

Comparative studies for evaluation of CO2 fixation in the cavity of the Rubisco enzyme using QM, QM/MM and linear-scaling DFT methods

Comparative studies for evaluation of CO2 fixation in the cavity of the Rubisco enzyme using QM, QM/MM and linear-scaling DFT methods

Linear-scaling density-functional simulations of charged point defects in Al2O3 using hierarchical sparse matrix algebra

Computing the Optical Properties of Large Systems

Contact Info

Product

Resources

About