Options
Physics-based and data-driven multiscale modeling of the structural formation in macromolecular systems
Citation Link: https://doi.org/10.15480/882.13251
Publikationstyp
Doctoral Thesis
Date Issued
2024
Sprache
English
Author(s)
Depta, Philipp Nicolas
Advisor
Referee
Title Granting Institution
Technische Universität Hamburg
Place of Title Granting Institution
Hamburg
Examination Date
2024-01-19
TORE-DOI
First published in
Number in series
25
Citation
Cuvillier Verlag (ISSN 2943-8500, Bd. 25, 2024)
Publisher
Cuvillier Verlag
ISBN
978-3-7369-7972-7
978-3-7369-6972-8
Peer Reviewed
true
Macromolecular structural formation and hierarchical self-assembly is crucial for a variety of systems in both nature and technology. Such systems may retain a remarkable structural organization from the atomistic up to the macroscopic scale enabling crucial features for their function. Many of these systems achieve this through self-assembly and consequently do not rely on external assembly mechanisms. In the field of material science one example is hydrogels, which achieve significant mechanical strength through polymer network formation on the molecular level. In the field of biology examples are abundant including most enzymes and viruses. One example is the hepatitis B virus, which contains a structural protein that assembles into regular spherical structures to transport the genetic material of the virus. Another example is the pyruvate dehydrogenase complex, which is crucial for cellular respiration and achieves its high biocatalytic activity through structural formation, thereby enabling features such as metabolic channeling. While there is an abundant amount of examples, investigation is challenging both experimentally and numerically. The phenomena involved in such structural assembly spread over vast scales in length and time and contain not only regular structures, but often also disordered elements. Consequently, capturing the mechanisms of formation, especially their kinetics, is inherently difficult.
In order to improve understanding of these phenomena, this work proposes a physics-based and data-driven multiscale modeling framework capable of describing structural formation on the micro-meter and milli-second scale, while retaining large amounts of molecular detail. The framework achieves this by abstracting the elementary macromolecules of a system as anisotropic unit objects and describes the interaction between units as well as the environment through data-driven models, e.g. 6D interaction potential fields. The models are parameterized in a bottom-up fashion and validated top-down. The framework is applied to and validated on three model systems: the gelation of alginate in CaCl2 solution, the self-assembly of the hepatitis B core antigen into virus-like particles, and the assembly and agglomeration of the pyruvate dehydrogenase complex. Results are validated using literature data and experimental data provided by collaborators, which show good agreement with measurable characteristics. Consequently, the developed framework enables novel scales to be investigated using numerical simulations and proposes a streamlined bottom-up parameterization, thus paving the way towards physically-mechanistic modeling of such structural assembly processes.
In order to improve understanding of these phenomena, this work proposes a physics-based and data-driven multiscale modeling framework capable of describing structural formation on the micro-meter and milli-second scale, while retaining large amounts of molecular detail. The framework achieves this by abstracting the elementary macromolecules of a system as anisotropic unit objects and describes the interaction between units as well as the environment through data-driven models, e.g. 6D interaction potential fields. The models are parameterized in a bottom-up fashion and validated top-down. The framework is applied to and validated on three model systems: the gelation of alginate in CaCl2 solution, the self-assembly of the hepatitis B core antigen into virus-like particles, and the assembly and agglomeration of the pyruvate dehydrogenase complex. Results are validated using literature data and experimental data provided by collaborators, which show good agreement with measurable characteristics. Consequently, the developed framework enables novel scales to be investigated using numerical simulations and proposes a streamlined bottom-up parameterization, thus paving the way towards physically-mechanistic modeling of such structural assembly processes.
Subjects
multiscale modeling
molecular modeling
Molecular Discrete Element Method
MDEM
Discrete Element Method
DEM
coarse-graining
Molecular Dynamics
MD
Langevin dynamics
machine learning
ML
supervised learning
Kriging
macromolecular self-assembly
structural formation simulation
anisotropic macromolecules
assembly pathways
assembly kinetics
molecular collisions
6D intermolecular interaction potentials
specialized force-fields
molecular binding
bonded interaction
hepatitis B core antigen
HBcAg
capsid formation
virus-like particles
VLP
pyruvate dehydrogenase complex
PDC
alginate
alginic acid
biopolymer
gelation
gel
aerogel
porous nanomaterial
anisotropic diffusion
ion binding model
calcium
proteins
enzymes
multi-enzymatic biocatalysis
metabolic channeling
high performance computing
HLRS
GPU implementation
MUSEN
DDC Class
540: Chemistry
570: Life Sciences, Biology
620.1: Engineering Mechanics and Materials Science
Funding Organisations
More Funding Information
The author gratefully acknowledges the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG, HE 4526/19-2 within the priority program SPP 1934) and High-Performance Computing Center Stuttgart (HLRS, Acid 44178).
Publication version
publishedVersion
Loading...
Name
9783736969728_eBook-1_mai.pdf
Size
84.77 MB
Format
Adobe PDF