Coarse grained models for molecular dynamic simulations of liquid structure and protein folding and self-assembly have been the subject of decades of research efforts. Although such models enable probing into longer length and time scales, they are still limited in accuracy and scale poorly to thermodynamic conditions or chemical entities beyond the ones at which they are developed. In this work, I leverage a powerful coarse graining framework based on an information theoretic metric known as the relative entropy to design novel coarse grained models for equilibrium phase behavior in liquid mixtures that correctly address the relevant manybody physics critically involved in phase behavior and thus, can provide structurally accurate descriptions of macroscopic phase separation across a large range of mixture compositions. I also develop protein models that are “next- generation” in the sense that they are systematically extendable in complexity while depending minimally on experimental data. These protein models offer remarkable predictive accuracy for folding of single proteins and insights into large scale protein self-assembly commonly seen in neurodegenerative pathologies such as Alzheimer’s and prion diseases. This dissertation develops and applies powerful coarse-graining techniques to meet longstanding challenges in the field and elevate coarse grained models from being “toy” systems to accurate, “production-level” tools for the development of biotechnologies and advanced materials.

Understanding and manipulating the behavior of water is a crucial component of designing materials at interfaces (i.e. for use in membranes, filters, chromatographic columns, etc.) as well as materials based on self-assembling molecules (i.e. micelles, hydrogels, etc.). It has long been known that water plays a significant role in mediating surface adsorption as well as interactions between biomolecules, but effective strategies to modify this behavior based on atomistic physical intuition has been lacking. Many past efforts have focused on unrealistic, toy systems in order to develop an extensive theory of water behavior in idealized inhomogeneous environments. Here I put this theory to the test in realistic yet simple and well-controlled molecular simulations, beginning the process of identifying practical extensions to inform the design of real materials.

To isolate and understand key water physics, I investigate ever-important hydrophobic interactions and explore the thermodynamic information contained within water’s structural response to hydrophobes. Specifically, consideration of the behavior of three-body angle distributions, which relate to the tetrahedral structure of water, allows for the construction of simple thermodynamic models for adsorption and solvation. I show that, in general, the success of such thermodynamic models may be attributed to fundamental connections between perturbations to water structure and the relative entropy of a solvation process, which includes indirect thermodynamic contributions associated with solvent restructuring.

In proteins and other biological settings, hydrophobic regions do not occur in isolation, but instead are part of heterogeneous interfaces involving a variety of chemical groups. To study the role of chemical patterning in inducing unique water behavior at such surfaces, I present methods coupling a genetic algorithm to molecular dynamics simulations to optimize arbitrary surface properties through the reorganization of functional groups. Water mobility near an interface is optimized in this way, ultimately revealing novel relationships between water dynamics, structure, and thermodynamics. I also use this algorithm to discover interfacial chemical patterns that repel or attract a diverse set of small-molecule solutes, as might be necessary to prevent fouling during water purification processes or selectively capture desirable products during chromatography. Overall, I extend our current knowledge of water’s molecular-scale response to both solutes and macroscopic interfaces, effectively expanding the design space for interfacial materials in terms of these atomistic-level insights.

Rising risks of climate change and supply chain insecurity highlight the need to develop alternative, greener synthesis routes to common materials currently sourced from petroleum. Biological systems excel at interconverting chemicals with exquisite specificity and speed, using networks of enzymes that perform catalysis at mild conditions. Protein complexes in nature colocalize complementary subunits to perform sophisticated biochemistry, and artificial, spatial organization of enzyme systems into synthetic complexes is an attractive strategy for improving biocatalytic process throughputs in industrial settings. While some sets of modular parts that enable designer protein complex construction exist, there is still a need to develop new components that are widely compatible with different enzymes and that are highly engineerable to impart desired self-assembly properties. Fungal cellulosomes, modular protein machines produced by anaerobic fungi in the guts of herbivores to rapidly free sugars from plant matter, represent an unexplored framework for synthetic protein complex construction. Cellulosomes synergistically incorporate enzymes involved in biomass degradation into discrete complexes via modular protein-protein interactions between enzyme fused dockerin domains and cohesin domains repeated on a central scaffoldin protein. Over 80% of the degradative power anaerobic fungi possess is attributed to cellulosomes, but the mechanistic nature of their activity and their assembly mechanism remain unknown. These knowledge gaps have precluded the development of fungal cellulosomes or their parts as biocatalytic technologies with real world applications. We apply a range of experimental techniques towards addressing how cellulosomes are produced in native anaerobic fungal cultures and characterizing the composition, nanostructure, and biochemical activity of purified, native cellulosomes. Immunofluorescence microscopy with cellulosome-labeling antibodies shows cellulosomes localize to the surfaces of cells, but that only cells at certain stages of the multi-staged life cycle produce cellulosomes under specific growth conditions. A robust cellulosome purification method we developed, in conjunction with mass spectrometry-based proteomics and biomass hydrolysis kinetic assays, provides high resolution details into the composition and lignocellulolytic activities of isolated cellulosomes produced by an anaerobic fungus, advancing our understanding of how cellulosomes can be engineered to enhance biomass hydrolysis rates. Towards leveraging the modular cellulosome assembly framework for synthetic biology applications, we develop a suite of modular interacting parts for constructing protein complexes with fungal cellulosome proteins. Through a combination of molecular modeling and high-throughput screening, we engineer interacting domains with a range of pH dependent binding behaviors for building protein complexes whose composition and therefore function are modulated with and environmental trigger, pH. Together, these tools and insights shed light on how cellulosomes make anaerobic fungi prolific biomass degraders and provide a framework for engineering protein complexes inspired by fungal cellulosomes designed for a wide range of applications.

Understanding the solution behavior of complex soft materials is crucial for designing and optimizing formulations that are relevant in everyday consumer products, including processed foods, detergents, hair care products, and various industrial applications such as lubricants, pesticides, and coatings. These formulations are highly multi-component and involve a wide range of charged molecules, such as polyelectrolytes, surfactants, and colloids, often in the presence of salt and other non-ionic (macro)molecules. While experimental investigations provide valuable insights, they are often limited in their ability to directly observe molecular-level interactions and explore the vast design space, encompassing numerous parameters such as composition, specific chemical species, macromolecule architecture, molecular weight, temperature, pH, and more. Computational simulations offer a powerful tool to complement experimental studies, providing a high-throughput screening approach to deepen our understanding of the underlying molecular interactions and the behavior of complex formulations.

In this thesis, we present a multi-scale simulation approach that parameterizes mesoscopic models of the field theory based on information obtained from small-scale atomistic simulations. We employ the relative entropy minimization framework to derive chemically-sensitive coarse-grained interaction parameters from all-atom simulations. Subsequently, we utilize the exact transformation to convert the coarse-grained particle-based model into field-theoretic form, facilitating the prediction of solution phase behavior. The overall workflow preserves the chemical specificity in complex mixtures of interest, enabling de novo studies of solution phase behavior in the field theory without the need for any experimental input.

The simulation framework is highly adaptable and can be applied to investigate a wide range of soft-matter formulations. This thesis focuses on formulations that rely on the complexation of charged macromolecules. The presence of charged assemblies, such as micelles, in typical formulations introduces further complexity, including long-length and time-scale phenomena that are intractable with other high-resolution simulation techniques. Through the exploration of various complex formulations, we demonstrate the predictive capability of this simulation workflow in exploring the thermodynamics and complex structures arising in such formulations. By integrating atomistic and mesoscopic simulation techniques, this work contributes to a fundamental understanding of the underlying mechanisms governing solution behavior in complex formulations. It offers valuable insights for the rational design and optimization of soft matter formulations, thereby contributing to advancements in various industries and sustainable chemical product development.

When a charged surface, such as an electrode, colloid, or protein, is submerged into an electrolyte or ionic liquid, ions within the fluid rearrange into electric double layers (EDLs) that electrostatically screen the interfacial charge. The electrostatic potential and ion distributions within EDLs have long been described by mean-field local-density approximations (LDAs) that assume flat electrodes, uncorrelated ions, and bulk forms for the chemical potential. The objective of this work is to elucidate LDA failure mechanisms and supplement or supplant mean-field treatments of electrochemical systems that fail to capture correlated behavior. We develop an exceedingly general method, which requires no a priori model and identifies whether EDLs in a given electrolyte can obey a LDA, or whether more advanced approaches (e.g. integro-differential equations, atomistic simulations, etc.) are required, irrespective of the source of LDA breakdown. We combine continuum-level theoretical studies with complementary simulations in order to critically assess the accuracy of LDA models of implicit solvent electrolytes with equal and differently sized ions. We also pose a novel LDA model that seeks to address solvation, polarizability, and finite-size interactions present in actual and simulated EDLs with explicit solvent.