Computer models are used extensively to investigate biological systems. Many of these systems can be described in terms of fields???spatially- and temporally- varying scalar, vector and tensor properties defined over domains. For example, the spatial variation of muscle fibers is a vector field, the spatial and temporal variation in temperature of an organ is a scalar field, and the distribution of stress across muscle tissue is a tensor field. In this thesis I present my research on how to represent fields in a format that allows researchers to store and distribute them independently of models and to investigate and manipulate them intuitively. I also demonstrate how the work can be applied to solving and analysing biomedical models. To represent fields I created a two-layer system. One layer, called the Field Representation Language (FRL), represents fields by storing numeric, analytic and meta data for storage and distribution. The focus of this layer is efficiency rather than usability. The second layer, called the Abstract Field Layer (AFL), provides an abstraction of fields so that they are easier for researchers to work with. This layer also provides common operations for manipulating fields as well as transparent conversion to and from FRL representations. The applications that I used to demonstrate the use of AFL and FRL are (a) a fields visualisation toolkit, (b) integration of models from different scales and solvers, and (c) a solver that uses AFL internally. The layered architecture facilitated the development of tools that use fields. A similar architecture may also prove useful for representations of other modelled entities.

Thesis (M.S.)--University of North Carolina at Wilmington, 2006. / Includes bibliographical references (leaf: 26)

Pulse synchronization is the evolution of spontaneous firing action across a network of sensor nodes. In the pulse synchronization model all nodes across a network produce a pulse, or "fire", at regular intervals even without access to a shared global time. Previous researchers have proposed the Reachback Firefly algorithm for pulse synchronization, in which nodes react to the firings of other nodes by changing their period. We propose an extension to this algorithm for tolerating arbitrary or Byzantine faults of nodes. Our algorithm queues up all the firings heard in the current cycle and discards outliers at the end of the cycle. An adjustment is computed with the remaining values and used as a starting point of the next cycle. Through simulation we validate the performance of our algorithm and study the overhead in terms of convergence time and periodicity. The simulation considers two specific kinds of Byzantine faults, the No Jump model where faulty nodes follow their own firing cycle without reacting to firings heard from other nodes and the Random Jump model where faulty nodes fire at any random time in their cycle.

Pulse synchronization is the evolution of spontaneous firing action across a network of sensor nodes. In the pulse synchronization model all nodes across a network produce a pulse, or "fire", at regular intervals even without access to a shared global time. Previous researchers have proposed the Reachback Firefly algorithm for pulse synchronization, in which nodes react to the firings of other nodes by changing their period. We propose an extension to this algorithm for tolerating arbitrary or Byzantine faults of nodes. Our algorithm queues up all the firings heard in the current cycle and discards outliers at the end of the cycle. An adjustment is computed with the remaining values and used as a starting point of the next cycle. Through simulation we validate the performance of our algorithm and study the overhead in terms of convergence time and periodicity. The simulation considers two specific kinds of Byzantine faults, the No Jump model where faulty nodes follow their own firing cycle without reacting to firings heard from other nodes and the Random Jump model where faulty nodes fire at any random time in their cycle.

A model of an artificial ecosystem has been formulated for use as a tool to investigate the dynamics of autonomous biosystems. The model is part of a composite model of an EcoCyborg which consists of an ecosystem and its control system, both of which are contained inside a cylindrical space station. The objectives of this project were to design a model of the ecosystem, and to develop a method for its creation and implementation within the overall framework of the EcoCyborg Project. / The modeling approach that has been adopted for the ecosystem model is individual-based and object-oriented. This enables the inclusion of a description of the abiotic environment, as well as of the organisms that inhabit it. A total of 1000 species representing a range of taxonomic groups may be modeled. Individuals in each species are described by their behaviours and phenotypic traits. / The ecosystem model will be linked with the other components of the EcoCyborg model in a multi-process simulation under OS/2 Warp. The behaviour of the system will be studied to elucidate preliminary guidelines for the design, maintenance and control of complex systems.

This thesis is concerned with the formation and nature of intravesicular solids in chemical and biological systems. The precipitation of Ag₂O within unilamellar vesicle microvolumes is described in detail. Formation of small (<10nm) single-domain cubic Ag₂O crystallites occurs on membrane diffusion of hydroxide ions. Nucleation initiates at a single site on the inner membrane surface followed by slow crystal growth. No precipitation is observed below an extravesicular pH (pH_OUT) of <u>ca</u>. 11.0. Permeable intravesicular nitrate ions permit hydroxide influx only when a critical membrane potential gradient is surpassed. Above a pH_OUT of 11.0 rate of precipitation is dependent on the rate of crystal growth. Kinetics are firstorder with respect to intravesicular Ag(I) concentration and approximately first-order with respect to hydroxide concentration below a pH_OUT of 12.0. Solubility equilibria for intra-and extravesicular Ag₂O formation are the same. Further <u>in situ</u> intravesicular precipitation reactions are described for FeO(OH) (crystalline), FeO(OH) (amorphous), Ag₂SiO₃ (crystalline), CoSiO₃ (amorphous), Co(OH)₂ (amorphous), Ag₂S (crystalline), CoS (amorphous), and Agl (crystalline). Precipitation of single oxidation state compounds is the same in vesicle space as for normal aqueous solution. For a mixed valency state solid, Fe₃0₄, intravesicular precipitation results in an amorphous material compared with a crystalline material prepared in aqueous solution. Inclusion of pre-formed Fe₃0₄, into vesicles is of potential use as a magnetic drug carrier system and n.m.r. relaxation probe. Intravesicular silica deposition in <u>Stephenaoeca diplocostata</u> Ellis is investigated. Siliceous costal strips are found to be extremely amorphous in structure, have surfaces active to Co(II) and Fe(III) ions, and demineralise from their centre of axis. T-joins of costal strips in intact loricae are found to be joined by a connective material containing amorphous silica centred around a filamentous material of unknown composition. Sectioned material indicates that silicification possibly initiates on an organic preformer laid down within an elongated intracellular vesicle. Potential EM stains, K₅SiPhGeW₁₁0₃₉, Gd(fod)₃, and t-BuNH₃[(nC₂₂H₄₅PO₃)₂Mo₅O₁₅] interact with vesicle bilayers resulting in observable images in the EM. Binding of K₅SiPhGeW₁₁0₃₉ can be followed by ¹H n.m.r. spectroscopy . Gd(fod)₃ loaded vesicles are potential n.m.r. probes for protein - membrane binding studies.

<i>Saccharomyces cerevisiae </i>is able to sense and respond to amino acid availability in the environment by reprogramming its gene expression.  Upon starvation for any amino acid, the yeast cell induces the biosynthesis of nearly all amino acids by upregulating the master transcription factor Gcn4, which is primarily controlled at the translational level.  Amino acid consumption is decreased by reducing general mRNA translation activity via <i>GCN2.  </i>The molecular mechanisms that mediate this response are known as the General Control Nonderepressible (GCN) response.  The aim of this thesis was to quantitatively understand the GCN response in <i>S. cerevisiae </i>by constructing a series of mathematical models, both deterministic and stochastic, based on published experimental results. Firstly, a deterministic model of general mRNA translation is described, which elaborates the translation initiation apparatus.  This model was used to predict the effects upon general translation activity of changing the abundance of specific initiation factors in isolation and collectively.  This model was also used to study the robustness of general mRNA translation and the trade-off between robustness and performance of this system. Secondly, a series of probabilistic and stochastic models of <i>GCN4 </i>mRNA translation based on Gillespie algorithm are described.  These probabilistic models successfully explained published results regarding the changes in <i>GCN4 </i>mRNA translation resulting from variation in uORF intercistronic distances.  The subsequent stochastic simulations suggested that histidine codon translation rates contribute significantly to the observed changes in ribosomal loading that occur on the <i>GCN4 </i>mRNA. Finally, a comprehensive deterministic model is described for the entire GCN system, integrating the above models.  This comprehensive model was able to predict the GCN responses to both natural and artificial amino acid starvation.  It successfully reproduced the phenotypes of some <i>GCN2 </i>and <i>GCN4 </i>mutants, and was also used to examine the fragility of GCN system when faced with severe artificial histidine starvation.  These predictions await experimental verification.

In this thesis, a philosophy and lexicon for the engineering of biosystems are established. The focus is on a specific class of biosystems ( ecocyborgs) created by combining ecosystems and technological Components. This work is part of the EcoCyborg Project, a highly interdisciplinary research program which concerns the development of a general theory for biosystems engineering, with an emphasis on system autonomy as a design goal. In the short term, the objective is to develop computational models and simulations for use in the study of ecocyborgs as representative instances of substantially autonomous biosystems. Accordingly, in this thesis an explicit conceptual basis is established for the EcoCyborg Project, as well as for biosystems engineering in general. / First, in the body of the thesis, a biosystem is defined as a coherent assemblage of entities that is alive to some degree as a whole. The sole criterion for life is considered to be comportment that is somewhat autopoietic , whereby local interactions among the components combine to Continually renew the overall system. Next, concepts related to autonomy, or the formulation and pursuit of proprietary goals, are elaborated. The degree of autonomy of a system is seen to depend on its consciousness, or ability to reason using a model of itself. Hence, a substantially autonomous system requires an ensemble of information storage and processing devices (mind) of the type and sophistication (intelligence ) appropriate for this. The approach that is taken here to the creation of ecocyborgs with such minds is described, and a specific mental architecture is delineated, comprising functionally semidifferentiated, intermediate-scale components arranged according to a semihierarchical control organization. Finally, the characterization of such systems is scrutinized as an epistemic process in which knowledge is generated by an observer, but in which only a limited degree of objectivity is possible. A paradigm appropriate to the engineering of ecocyborgs is defined as an illustration, and associated archetypal concepts and descriptive procedures (such as measures) are given that are useful in this context. Such tools are required by significantly autonomous ecocyborgs because they must characterize themselves. They are also necessary to observers with scientific and engineering agendas.

A Systematic Approach to Bio-inspired Conceptual Design

Computer models are used extensively to investigate biological systems. Many of these systems can be described in terms of fields???spatially- and temporally- varying scalar, vector and tensor properties defined over domains. For example, the spatial variation of muscle fibers is a vector field, the spatial and temporal variation in temperature of an organ is a scalar field, and the distribution of stress across muscle tissue is a tensor field. In this thesis I present my research on how to represent fields in a format that allows researchers to store and distribute them independently of models and to investigate and manipulate them intuitively. I also demonstrate how the work can be applied to solving and analysing biomedical models. To represent fields I created a two-layer system. One layer, called the Field Representation Language (FRL), represents fields by storing numeric, analytic and meta data for storage and distribution. The focus of this layer is efficiency rather than usability. The second layer, called the Abstract Field Layer (AFL), provides an abstraction of fields so that they are easier for researchers to work with. This layer also provides common operations for manipulating fields as well as transparent conversion to and from FRL representations. The applications that I used to demonstrate the use of AFL and FRL are (a) a fields visualisation toolkit, (b) integration of models from different scales and solvers, and (c) a solver that uses AFL internally. The layered architecture facilitated the development of tools that use fields. A similar architecture may also prove useful for representations of other modelled entities.