arXiv subjects

Wenfei Li

Publications and source records attributed to Wenfei Li.

16 recordsLinked to original sources

arXiv2026

StochasticGW-GPU: rapid quasi-particle energies for molecules beyond 10000 atoms

$\mathtt{StochasticGW}$ is a code for computing accurate Quasi-Particle (QP) energies of molecules and material systems in the GW approximation. $\mathtt{StochasticGW}$ utilizes the stochastic Resolution of the Identity (sROI) technique to enable a massively-parallel implementation with computational costs that scale semi-linearly with system size, allowing the method to access systems with tens of thousands of electrons. We introduce a new implementation, $\mathtt{StochasticGW-GPU}$, for which the main bottleneck steps have been ported to GPUs and which gives substantial performance improvements over previous versions of the code. We showcase the new code by computing band gaps of hydrogenated silicon clusters ($\textrm{S}\textrm{i}_{\textrm{x}}\textrm{H}_{\textrm{y}}$) containing up to 10001 atoms and 35144 electrons, and we obtain individual QP energies with a statistical precision of better than $\pm0.03$ eV with times-to-solution on the order of minutes.

physics.chem-ph↗

arXiv2023

DeePKS Model for Halide Perovskites with the Accuracy of Hybrid Functional

Accurate prediction for the electronic structure properties of halide perovskites plays a significant role in the design of highly efficient and stable solar cells. While density functional theory (DFT) within the generalized gradient approximation (GGA) offers reliable prediction in terms of lattice constants and potential energy surface for halide perovskites, it severely underestimates the band gap due to the lack of non-local exact exchange term, which exists in computationally expensive hybrid functionals. In this work, a universal Deep Kohn-Sham (DeePKS) model based on neural network is trained so as to enable electronic structure calculations with the accuracy of hybrid functional HSE06 and the efficiency comparable to GGA functional, for a plethora of halide perovskites, i.e., ABX$_3$ (A=FA, MA, Cs; B=Sn, Pb; X=Cl, Br, I). Forces, band gaps, and density of states (DOS) predicted by our DeePKS model for all aforementioned perovskites are in good agreement with the HSE06 results, with significantly improved efficiency. In addition, even though the spin-orbit coupling (SOC) effect has not been taken into consideration during the training process, DeePKS+SOC offers highly consistent band gap and DOS as compared to HSE06+SOC for Pb-containing systems. We believe such DeePKS model can be readily applied for an accurate yet efficient prediction of various properties for the family of halide perovskites.

physics.chem-ph↗

arXiv2023

Implementation of the SCAN Exchange-Correlation Functional with Numerical Atomic Orbitals

Kohn-Sham density functional theory (DFT) is nowadays widely used for electronic structure theory simulations, and the accuracy and efficiency of DFT rely on approximations of the exchange-correlation functional. By inclusion of the kinetic energy density $τ$, the meta-generalized-gradient approximation (meta-GGA) family of functionals achieves better accuracy and flexibility while retaining the efficiency of semi-local functionals. The SCAN meta-GGA functional has been proven to yield accurate results for solid and molecular systems. We implement meta-GGA functionals with both numerical atomic orbitals and plane wave basis in the ABACUS package. Apart from the exchange-correlation potential, we also discuss the evaluation of force and stress. To validate our implementation, we perform finite-difference tests and convergence tests with the SCAN meta-GGA functional. We further test water hexamers, weakly interacting molecules of the S22 dataset, as well as 13 semiconductors. The results show satisfactory agreements with previous calculations and available experimental values.

cond-mat.mtrl-sci↗

arXiv2023

Bidirectional allostery mechanism of catch-bond effect in cell adhesion

Catch-bonds, whereby noncovalent ligand-receptor interactions are counterintuitively reinforced by tensile forces, play a major role in cell adhesion under mechanical stress. A basic prerequisite for catch-bond formation is that force-induced remodeling of ligand binding interface occurs prior to bond rupture. However, what strategy receptor proteins utilize to meet such specific kinetic control is still unclear, rendering the mechanistic understanding of catch-bond an open question. Here we report a bidirectional allostery mechanism of catch-bond for the hyaluronan (HA) receptor CD44 which is responsible for rolling adhesion of lymphocytes and circulating tumor cells. Binding of ligand HA allosterically reduces the threshold force for unlocking of otherwise stably folded force-sensing element (i.e., forward allostery), so that much smaller tensile force can trigger the conformational switching of receptor protein to high binding-strength state via backward allosteric coupling before bond rupture. The effect of forward allostery was further supported by performing atomistic molecular dynamics simulations. Such bidirectional allostery mechanism fulfills the specific kinetic control required by catch-bond and is likely to be commonly utilized in cell adhesion. We also revealed a slip-catch-slip triphasic pattern in force response of CD44-HA bond arising from force-induced repartitioning of parallel dissociation pathways. The essential thermodynamic and kinetic features of receptor proteins for shaping the catch-bond were identified.

physics.bio-ph↗

arXiv2022

DeePKS+ABACUS as a Bridge between Expensive Quantum Mechanical Models and Machine Learning Potentials

Recently, the development of machine learning (ML) potentials has made it possible to perform large-scale and long-time molecular simulations with the accuracy of quantum mechanical (QM) models. However, for high-level QM methods, such as density functional theory (DFT) at the meta-GGA level and/or with exact exchange, quantum Monte Carlo, etc., generating a sufficient amount of data for training a ML potential has remained computationally challenging due to their high cost. In this work, we demonstrate that this issue can be largely alleviated with Deep Kohn-Sham (DeePKS), a ML-based DFT model. DeePKS employs a computationally efficient neural network-based functional model to construct a correction term added upon a cheap DFT model. Upon training, DeePKS offers closely-matched energies and forces compared with high-level QM method, but the number of training data required is orders of magnitude less than that required for training a reliable ML potential. As such, DeePKS can serve as a bridge between expensive QM models and ML potentials: one can generate a decent amount of high-accuracy QM data to train a DeePKS model, and then use the DeePKS model to label a much larger amount of configurations to train a ML potential. This scheme for periodic systems is implemented in a DFT package ABACUS, which is open-source and ready for use in various applications.

physics.chem-ph↗

arXiv2021

Tempering stochastic density functional theory

We introduce a tempering approach with stochastic density functional theory (sDFT), labeled t-sDFT, which reduces the statistical errors in the estimates of observable expectation values. This is achieved by rewriting the electronic density as a sum of a "warm" component complemented by "colder" correction(s). Since the "warm" component is larger in magnitude but faster to evaluate, we use many more stochastic orbitals for its evaluation than for the smaller-sized colder correction(s). This results in a significant reduction of the statistical fluctuations and the bias compared to sDFT for the same computational effort. We the method's performance on large hydrogen-passivated silicon nanocrystals (NCs), finding a reduction in the systematic error in the energy by more than an order of magnitude, while the systematic errors in the forces are also quenched. Similarly, the statistical fluctuations are reduced by factors of around 4-5 for the total energy and around 1.5-2 for the forces on the atoms. Since the embedding in t-sDFT is fully stochastic, it is possible to combine t-sDFT with other variants of sDFT such as energy-window sDFT and embedded-fragmented sDFT.

physics.comp-ph↗

arXiv2021

Tuning the range separation parameter in periodic systems

Kohn-Sham DFT with optimally tuned range-separated hybrid (RSH) functionals provides accurate and nonempirical fundamental gaps for a wide variety of finite-size systems. The standard tuning procedure relies on calculation of total energies of charged systems and thus cannot be applied to periodic solids. Here, we develop a framework for tuning the range separation parameter that can be used for periodic and open boundary conditions. The basic idea is to choose the range parameter that results in a stationary point where the fundamental gap obtained by RSH matches the gap obtained from a G0W0 over RSH calculation. The proposed framework is therefore analogous to eigenvalue self-consistent GW (scGW). We assess the method for various solids and obtain very good agreement with scGW results.

cond-mat.mtrl-sci↗

arXiv2020

Real Space Orthogonal Projector-Augmented-Wave Method

The projector augmented wave (PAW) method of Blöchl makes smooth but non-orthogonal orbitals. Here we show how to make PAW orthogonal, using a cheap transformation of the wave-functions. We show that the resulting Orthogonal PAW (OPAW), applied for DFT, reproduces (for a large variety of solids) band gaps from the ABINIT package. OPAW combines the underlying orthogonality of norm-conserving potentials with the large grid spacings and small spectral range in PAW. The OPAW framework can also be combined with other electronic structure theory methods.

physics.comp-ph↗

arXiv2020

Computational prediction of RNA tertiary structures using machine learning methods

RNAs play crucial and versatile roles in biological processes. Computational prediction approaches can help to understand RNA structures and their stabilizing factors, thus providing information on their functions, and facilitating the design of new RNAs. Machine learning (ML) techniques have made tremendous progress in many fields in the past few years. Although their usage in protein-related fields has a long history, the use of ML methods in predicting RNA tertiary structures is new and rare. Here, we review the recent advances of using ML methods on RNA structure predictions and discuss the advantages and limitation, the difficulties and potentials of these approaches when applied in the field.

physics.bio-ph↗

arXiv2020

Dopant levels in large nanocrystals using stochastic optimally tuned range-separated hybrid density functional theory

We apply a stochastic version of an optimally tuned range-separated hybrid functional to provide insight on the electronic properties of P- and B- doped Si nanocrystals of experimentally relevant sizes. We show that we can use the range-separation parameter for undoped systems to calculate accurate results for dopant activation energies. We apply this strategy for tuning functionals to study doped nanocrystals up to 2.5 nm in diameter at the hybrid functional level. In this confinement regime, the P- and B- dopants have large activation energies and have strongly localized states that lie deep within the energy gaps. Structural relaxation plays a greater role for B-substituted dopants and contributes to the increase in activation energy when the B dopant is near the nanocrystal surface.

cond-mat.mtrl-sci↗

arXiv2019

Stochastic embedding DFT: theory and application to p-nitroaniline

Over this past decade, we combined the idea of stochastic resolution of identity with a variety of electronic structure methods. In our stochastic Kohn-Sham DFT method, the density is an average over multiple stochastic samples, with stochastic errors that decrease as the inverse square root of the number of sampling orbitals. Here we develop a stochastic embedding density functional theory method (se-DFT) that selectively reduces the stochastic error (specifically on the forces) for a selected sub-system(s). The motivation, similar to that of other quantum embedding methods, is that for many systems of practical interest the properties are often determined by only a small sub-system. In stochastic embedding DFT two sets of orbitals are used: a deterministic one associated with the embedded subspace, and the rest which is described by a stochastic set. The method is exact in the limit of large number of stochastic samples. We apply se-DFT to study a p-nitroaniline molecule in water, where the statistical errors in the forces on the system (the p-nitroaniline molecule) are reduced by an order of magnitude compared with non-embedding stochastic DFT.

physics.chem-ph↗

arXiv2018

Swift $GW$ beyond $10,000$ electrons using fractured stochastic orbitals

We introduce the concept of fractured stochastic orbitals (FSOs), short vectors that sample a small number of space points and enable an efficient stochastic sampling of any general function. As a first demonstration, FSOs are applied in conjunction with simple direct-projection to accelerate our recent stochastic $GW$ technique; the new developments enable accurate prediction of $G_{0}W_{0}$ quasiparticle energies and gaps for systems with up to $N_{e}>10,000$ electrons, with small statistical errors of $\pm0.05\,{\rm eV}$ and using less than 2000 core CPU hours. Overall, stochastic $GW$ scales now linearly (and often sub-linearly) with $N_{e}.$

physics.comp-ph↗

arXiv2009

Self-learning Multiscale Simulation for Achieving High Accuracy and High Efficiency Simultaneously

We propose a new multi-scale molecular dynamics simulation method which can achieve high accuracy and high sampling efficiency simultaneously without aforehand knowledge of the coarse grained (CG) potential and test it for a biomolecular system. Based on the resolution exchange simulations between atomistic and CG replicas, a self-learning strategy is introduced to progressively improve the CG potential by an iterative way. Two tests show that, the new method can rapidly improve the CG potential and achieve efficient sampling even starting from an unrealistic CG potential. The resulting free energy agreed well with exact result and the convergence by the method was much faster than that by the replica exchange method. The method is generic and can be applied to many biological as well as non-biological problems.

physics.bio-ph↗

arXiv2008

Metal Coupled Folding of Cys2His2 Zinc-Finger

Zinc-fingers, which widely exist in eukaryotic cell and play crucial roles in life processes, depend on the binding of zinc ion for their proper folding. To computationally study the zinc coupled folding of the zinc-fingers, charge transfer and metal induced protonation/deprotonation effects have to be considered. Here, by attempting to implicitly account for such effects in classical molecular dynamics and performing intensive simulations with explicit solvent for the peptides with and without zinc binding, we investigate the folding of the Cys2His2 type zinc-finger motif and the coupling between the peptide folding and zinc binding. We find that zinc ion not only stabilizes the native structure, but also participates in the whole folding process. It binds to the peptide at early stage of folding, and directs or modulates the folding and stabilizations of the component beta-hairpin and alpha-helix. Such a crucial role of zinc binding is mediated by the packing of the conserved hydrophobic residues. We also find that the packing of the hydrophobic residues and the coordination of the native ligands are coupled. Meanwhile, the processes of zinc binding, mis-ligation, ligand exchange and zinc induced secondary structure conversion, as well as the water behaviour, due to the involvement of zinc ion are characterized. Our results are in good agreement with related experimental observations, and provide significant insight into the general mechanisms of the metal-cofactor dependent protein folding and other metal induced conformational changes of biological importance.

physics.bio-ph↗

arXiv2005

Particle transfer and fusion cross-section for Super-heavy nuclei in dinuclear system

Within the dinuclear system (DNS) conception, instead of solving Fokker-Planck Equation (FPE) analytically, the Master equation is solved numerically to calculate the fusion probability of super-heavy nuclei, so that the harmonic oscillator approximation to the potential energy of the DNS is avoided. The relative motion concerning the energy, the angular momentum, and the fragment deformation relaxations is explicitly treated to couple with the diffusion process, so that the nucleon transition probabilities, which are derived microscopically, are time-dependent. Comparing with the analytical solution of FPE, our results preserve more dynamical effects. The calculated evaporation residue cross sections for one-neutron emission channel of Pb-based reactions are basically in agreement with the known experimental data within one order of magnitude.

nucl-th↗

arXiv2005

Deformation and orientation effects in the driving potential of the dinuclear model

A double-folding method is used to calculate the nuclear and Coulomb interaction between two deformed nuclei with arbitrary orientations. A simplified Skryme-type interaction is adopted. The contributions of nuclear interaction and Coulomb interaction due to the deformation and orientation of the nuclei are evaluated for the driving potential used in the description of heavy-ion fusion reaction. So far there is no satisfactory theory to describe the evolution of the dynamical nuclear deformation and orientations during the heavy-ion fusion process. Our results estimated the magnitude of above effects.

nucl-th↗