Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Bowling_Thesis.pdf (28.73 MB)

ETD Abstract Container

Abstract Header

Quantum Mechanical Approaches for Large Protein Systems: Fragmentation, Confining Potentials, and Anisotropic Solvation

Bowling, Paige Elise
http://orcid.org/0000-0001-7709-9350
http://rave.ohiolink.edu/etdc/view?acc_num=osu1732208213250382

Abstract Details

2024, Doctor of Philosophy, Ohio State University, Biophysics.
Fragment-based quantum chemistry methods provide a way to circumvent the steep nonlinear scaling of electronic structure calculations, enabling the investigation of large molecular systems using high-level methods. First, we present calculations on enzyme models containing 500-600 atoms using the many-body expansion (MBE) and compare them to benchmarks where the entire enzyme-substrate complex is described at the same level of density functional theory (DFT). When amino acid fragments contain ionic side chains, the MBE exhibits oscillatory behavior under vacuum boundary conditions, but rapid convergence is restored using low-dielectric boundary conditions. This suggests that full-system gas-phase calculations are unsuitable as benchmarks for assessing errors in fragment-based approximations. A three-body protocol maintains sub-kcal/mol accuracy compared to supersystem calculations, as does a two-body approach combined with a low-cost full-system correction. In the next section, we use fragmentation to compute protein–ligand interaction energies in systems with several thousand atoms. Convergence tests using a minimal-basis semi-empirical method (HF-3c) indicate that two-body calculations, with single-residue fragments and simple hydrogen caps, are sufficient to reproduce interaction energies obtained using conventional supramolecular electronic structure calculations, to with 1 kcal/mol at about 1% of the cost. Additionally, we show that semi-empirical methods can be used as an alternative to DFT, to assess convergence of sequences of quantum mechanics (QM) models (of increasing size) generated by different automated protocols. Two-body calculations afford a low-cost way to construct a “QM-informed” enzyme model. This streamlined, user-friendly approach to building ligand binding-site models requires no prior information or manual adjustments, making it accessible and practical for a wide range of applications. For the latter parts of this work, we will be focusing on methods we have developed to improve the modeling of protein systems. Chapter 5 introduces methods to eliminate imaginary vibrational frequencies, which arise from artificial constraints in truncated active site models and prevent finite-temperature free energy corrections. We compare two approaches: replacing fixed-atom constraints with harmonic confining potentials and omitting fixed-atom contributions to the Hessian. While the latter eliminates imaginary frequencies, it underestimates zero-point energy and vibrational entropy, introducing artificial rigidity. Harmonic confining potentials eliminate imaginary frequencies and provide a flexible means to construct active-site models that can be used in unconstrained geometry relaxations, affording better convergence of reaction energies and barrier heights with respect to model size, as compared to models with fixed-atom constraints. The final chapter explores simulating anisotropic solvation, such as at interfaces or solvent-exposed regions in nonpolar environments. In this work, we have introduced a “heterogeneous” polarizable continuum model (PCM), where a dielectric constant is assigned separately to each atomic sphere that contributes to the solute cavity. For systems with large dielectric differences, this method diverges from traditional PCMs but aligns well with rigorous Poisson boundary conditions for smaller dielectric differences, yielding solvation energies within 2 kcal/mol. Applying this model to copper-containing proteins, the heterogeneous PCM more accurately predicts solvation free energies and pKa values compared to gas phase and homogeneous models, aligning closely with experimental data.
John Herbert (Advisor)
Sherwin Singer (Committee Member)
William Ray (Committee Member)
291 p.
Biochemistry; Biology; Biomedical Research; Biophysics; Chemistry; Computer Science; Molecular Biology; Molecular Chemistry; Molecular Physics; Molecules; Physical Chemistry; Physics; Quantum Physics; Technology; Theoretical Physics
Quantum Mechanics; Protein Thermochemistry; Method Development; Computational Biochemistry; Protein-Ligand Interaction Energies; Drug Discovery; Biophysics; Software Development; Fragmentation; Confining Potentials; Anisotropic Solvation

Recommended Citations

Citations

  • Bowling, P. E. (2024). Quantum Mechanical Approaches for Large Protein Systems: Fragmentation, Confining Potentials, and Anisotropic Solvation [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1732208213250382

    APA Style (7th edition)

  • Bowling, Paige. Quantum Mechanical Approaches for Large Protein Systems: Fragmentation, Confining Potentials, and Anisotropic Solvation. 2024. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1732208213250382.

    MLA Style (8th edition)

  • Bowling, Paige. "Quantum Mechanical Approaches for Large Protein Systems: Fragmentation, Confining Potentials, and Anisotropic Solvation." Doctoral dissertation, Ohio State University, 2024. http://rave.ohiolink.edu/etdc/view?acc_num=osu1732208213250382

    Chicago Manual of Style (17th edition)

osu1732208213250382
70
© , all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.