Toward ab initio refinement of protein X-ray crystal structures: interpreting and correlating structural fluctuations

Recommending Hartree–Fock Theory with London-Dispersion and Basis-Set-Superposition Corrections for the Optimization or Quantum Refinement of Protein Structures

“…We optimized a fragment of an ensemble of eight lysozyme structures 69 using periodic boundary conditions according to a modified divide-and-conquer approach. The lysozyme ensemble is based on highly accurate X-ray crystallographic data, 68 and it therefore served as a reliable benchmark.…”

“…Falklof et al addressed the problem of conformational flexibility in 2VB1 in detail in 2012 and set up a correlated ensemble structure for future quantum-chemical optimizations. 69 Because of the triclinic symmetry of the crystal, the simplest choice for a realistic description of the conformational variations was to take an ensemble of eight chemically complete structures. The originally determined occupancies of all atoms were each set to a multiple of one-eighth, with all atoms in identified correlated fluctuations (e.g., all atoms in a residue) assigned the same occupancies.…”

“…Falklof et al made the important observation that, even though the structure was obtained at very high resolution, only 25% of the residues of 2VB1 are surrounded by a 5-Å region in which the atomic structure is unambiguously defined. 69 Therefore, it was advised that only those well-defined regions be optimized and the structure be left unchanged for the others. In our previous study, 110 we followed this recommendation, and we do the same herein.…”

“…For the sake of simplicity, we chose the AMBER force-field 136 with B3LYP 76,77 /6-31G* 78 atomic charges taken from previous studies. 38,69 In preliminary calculations, we also explored the possibility of electronic embedding in the ONIOM calculations, that is, allowing the MM charges to directly influence the QM Hamiltonian. However, we observed that the results deteriorated, and therefore, we continued with only the standard ONIOM approach (see SI).…”

“…Previously some of us developed methods for the construction of such a mapping and applied it to HEWL. 69 This created an ensemble, shown in Figure 1, of eight chemically reasonable, fully interconnected, global protein structures that generate the same scattering R factor as the original PDB-deposited uncorrelated structure. Herein, we use this structure to investigate QM methods for protein-structure optimization and for quantum refinement.…”