An Efficient Implementation of the Generalized Energy-Based Fragmentation Approach for General Large Molecules

Abstract

An efficient implementation of the generalized energy-based fragmentation (GEBF) approach (Li, W.; Li, S.; Jiang, Y. J Phys. Chem. A 2007, 111, 2193) for treating a wide range of large molecules is presented. In this implementation, the fragmentation process can be automatically done for a general molecule, with only some functional groups defined by users. A new and fast scheme is designed for the generation of various subsystems and the derivation of their coefficients. The newly implemented GEBF approach has been applied to several large molecules including proteins, nucleic acids, and supermolecules with fused aromatic rings. Test calculations within the Hartree-Fock (HF) and density functional theory (DFT) framework demonstrate that the GEBF approach can provide reasonably accurate ground-state energies and optimized structures, which are in good agreement with those from conventional HF or DFT calculations. The GEBF approach implemented in this work can now be employed by nonexpert users to compute energies, optimized structures, and some molecular properties at various ab initio levels for a broad range of large molecules on ordinary PC workstations.

Understanding the many-body expansion for large systems. I. Precision considerations

“…As such, distance-based cutoffs are unlikely to be a panacea for loss of precision, so the development of high-accuracy methods that do not require four-body terms is therefore desirable. One promising approach is to use larger, overlapping fragments, 5,6,29,36,43,[52][53][54][55][56][57] in conjunction with a generalized many-body expansion. 5,6,29 Such an approach can provide accurate results at the two-body level, albeit at increased cost per subsystem calculation.…”

Linear scaling explicitly correlated MP2-F12 and ONIOM methods for the long-range interactions of the nanoscale clusters in methanol aqueous solutions

“…53 Then, primitive subsystems are built by adding environmental fragments for each fragment with a threshold ζ and derivative subsystems are constructed to guarantee each n-fragment term only occurring once. In order to take the long-range electrostatic interaction and polarization effects between remote fragments into account, each subsystem is embedded in the field of background point charges generated by all atoms outside this subsystem.…”

“…48 Alternatively, combined quantum mechanics and molecular mechanics (QM/MM) MD simulations are recently used for the methanol-water systems. 49,50 In the recent years, we have developed an energybased fragmentation (EBF) approach, 51 and a subsequent generalized energy-based fragmentation (GEBF) approach, [52][53][54] for the calculations of large systems. Other closed fragmentation methods have also been developed by other groups, [55][56][57][58][59][60][61][62][63][64] which including systematic molecular fragmentation by Deev and co-workers, [55][56][57] molecular tailoring approach by Gadre and co-workers, [58][59][60] a fragmentation method by Bettens and co-workers, 61,62 electrostatically embedded many-body expansion method by Sorkin, Dahlke, and Truhlar, 63 and molecules-in-molecules method by Mayhall and Raghavachari, 64 more fragmentation approaches can be found in recent review by Gordon et al 65 The basic idea of the GEBF approach is that the energy and properties of a target system can be approximately obtained from the conventional calculations of a series of subsystems, which are embedded in the field of background point charges generated by all atoms outside this subsystem, 52 and can be constructed automatically.…”

“…Other closed fragmentation methods have also been developed by other groups, [55][56][57][58][59][60][61][62][63][64] which including systematic molecular fragmentation by Deev and co-workers, [55][56][57] molecular tailoring approach by Gadre and co-workers, [58][59][60] a fragmentation method by Bettens and co-workers, 61,62 electrostatically embedded many-body expansion method by Sorkin, Dahlke, and Truhlar, 63 and molecules-in-molecules method by Mayhall and Raghavachari, 64 more fragmentation approaches can be found in recent review by Gordon et al 65 The basic idea of the GEBF approach is that the energy and properties of a target system can be approximately obtained from the conventional calculations of a series of subsystems, which are embedded in the field of background point charges generated by all atoms outside this subsystem, 52 and can be constructed automatically. 53 The GEBF approach has been implemented for the energy, dipole moments, and static polarizability calculations, 52 geometry optimizations and vibrational frequencies, 66 vibrational circular dichroism spectroscopies, 67 combined quantum mechanics and molecular mechanics (QM/MM) calculations, 68,69 of large systems at various levels of quantum chemistry methods, including Hartree-Fock (HF), DFT, MP2, and coupled cluster (CC) singles and doubles (CCSD) methods. [52][53][54] It has been demonstrated to be quite accurate for the relative energies, stabilities, and intermolecular interactions of medium-size water clusters and large-size macromolecules at DFT, MP2, and CCSD levels.…”

“…53 The GEBF approach has been implemented for the energy, dipole moments, and static polarizability calculations, 52 geometry optimizations and vibrational frequencies, 66 vibrational circular dichroism spectroscopies, 67 combined quantum mechanics and molecular mechanics (QM/MM) calculations, 68,69 of large systems at various levels of quantum chemistry methods, including Hartree-Fock (HF), DFT, MP2, and coupled cluster (CC) singles and doubles (CCSD) methods. [52][53][54] It has been demonstrated to be quite accurate for the relative energies, stabilities, and intermolecular interactions of medium-size water clusters and large-size macromolecules at DFT, MP2, and CCSD levels. 54,[70][71][72] As we know, the basis set limit of the conventional MP2 and CCSD correlation energies is difficult to be obtained due to the inaccurate description of short-range correlation with slater determinants in wavefunction.…”