Search

The laws of energy and material conservation are fundamental principles across various scales and systems. Based on the conservation laws, I derive several theoretical models to understand mechanisms behind the energy budget of ontogenetic growth and the pattern of the late Pleistocene extinction of megafauna in the Americas. First, I present a model, empirically grounded in data from birds and mammals, that correctly predicts how growing animals allocate food energy between synthesis of new biomass and maintenance of existing biomass. Previous energy budget models have typically been based on rates of either food consumption or metabolic energy expenditure. The model provides a framework that reconciles these two approaches and highlights the fundamental principles that determine rates of food assimilation and rates of energy allocation to maintenance, biosynthesis, activity, and storage. The model predicts that growth and assimilation rates for all animals should cluster closely around two canonical curves. Second, the previous model, which focuses on endotherms, has been extended to understand effects of temperature on the energy budget of ontogenetic growth of ectotherms. A tendency for ectotherms to develop faster but mature at smaller body sizes in warmer environments has been studied for decades, and is called the temperature size rule (TSR). It can be explained by a simple model in which the rate of growth or biomass accumulation and the rate of development or differentiation have different temperature dependence. The model accounts for both TSR and the less frequently observed reverse-TSR, predicts the fraction of energy allocated to maintenance and synthesis over the course of development, and the temperature independent growth efficiency. It also predicts that less total energy is expended when developing at warmer temperatures for TSR and vice versa for reverse-TSR. It has important implications for effects of climate change on ectothermic animals and al

My dissertation centers around investigating big-picture questions related to understanding the consequences of metabolism and energetics on the evolution, ecology, and physiology of life. The evolutionary transitions from prokaryotes to unicellular eukaryotes to multicellular organisms were accompanied by major innovations in metabolic design. In my first chapter, I show that the scaling of metabolic rate, population growth rate, and production efficiency with body size have changed across these transitions. Metabolic rate scales with body mass superlinearly in prokaryotes, linearly in protists, and sublinearly in metazoans, so Kleibers 3/4 power scaling law does not apply universally across organisms. This means that major changes in metabolic processes during the early evolution of life overcame existing physical constraints, exploited new opportunities, and imposed new constraints on organism physiology. Surface areas of physiological structures of organisms impose fundamental constraints on metabolic rate. In my second chapter, I demonstrate that organisms have a variety of options for increasing the scaling of the area of their metabolic surfaces with body sizes. I develop models and examples illustrating the role of cell membrane elaborations, mitochondria, vacuoles, vesicles, inclusions, and shape-shifting in the architectural design, evolution, and ecology of unicellular microbes. I demonstrate how these surface-area scaling adaptations have played important roles in the evolution of major biological designs of cells and the physiological ecology of organisms. In my third and final chapter, I integrate and synthesize findings from the previous two chapters with important developments in geochemistry, microbiology, and astrobiology in order to identify the fundamental physical and biological dimensions that characterize a metabolic theory of ecology of microorganisms. These dimensions are thermodynamics, chemical kinetics, physiological harshness, cell size

Bayesian belief aggregation is the process of forming a consensus model from the probabilistic beliefs of multiple individuals. Preference aggregation attempts to find an optimal solution for a population considering each individual's beliefs, desires and objectives. Belief and preference aggregation approaches that form a single consensus average away any diversity in a population. In the process they may fail to uphold a set of mathematical properties for rational aggregation defined by social choice theorists. This dissertation introduces a new aggregation approach that maintains the diversity of a population and allows the competitive aspects of a situation to emerge, enabling game theoretic analysis in large populations of decision-makers. Each individual's beliefs and preferences are represented by a Bayesian network. Based on the result of inference on the networks, a population is separated into collectives whose members agree on the relatively likelihood or desirability of the possible outcomes of a situation. An aggregate for each collective can then be computed such that the aggregate upholds the rationality properties. Game theoretic analysis is then applied using 'super-agents' that represent each collective as the game players. In this manner, the set of Pareto optimal and Nash equilibrium solutions can be found, even in situations that cause single consensus models to return non-Pareto or otherwise 'irrational' solutions.

Spatial heterogeneity in the distribution of food is an important determinant of species' optimal foraging strategies, and of the dynamics of populations and communities. In order to explore the interaction of food heterogeneity and colony size in their effects on the behavior of foraging ant colonies, we built agent-based models of the foraging and recruitment behavior of harvester ants of the genus Pogonomyrmex. We optimized the behavior of these models using genetic algorithms over a variety of food distributions and colony sizes, and validated their behavior by comparison with data collected on harvester ants foraging for seeds in the field. We compared two models: one in which ants lay a pheromone trail each time they return to the nest with food; and another in which ants lay pheromone trails selectively, depending on the density of other food available in the area where food was found. We found that the density-dependent trail-laying model fit the field data better. We found that in this density-dependent recruitment model, colonies of all sizes evolved intense recruitment behavior, even when optimized for environments in which the majority of foods are distributed homogeneously. We discuss the implications of these models to the understanding of optimal foraging strategy and community dynamics among ants, and potential for application to ACO and other distributed problem-solving systems.

The dengue virus is the causative agent of an important re-emerging infectious disease that has become increasingly significant in tropical America and the Caribbean due to the infiltration of a more pathogenic Asian/American strain of dengue serotype 2 into the population. This invading strain is responsible for epidemics of dengue hemorrhagic fever, a life-threatening disease that was not previously a large public health concern in the region. Here, I create a historical map of the invasion and replacement of the endemic American strain of dengue 2 by the Asian/American strain, showing that the timing of invasion spans 25 years, and is highly variable in the region. In addition, I model the competitive dynamics of the two strains using differential equa- tion models. By calculating and comparing the basic reproductive ratio (R0) for the Asian/American and American strain of dengue 2, I identify potential evolutionary trade-offs between the two strains and the ecological circumstances that benefit one trade-off over another. Numerically solving my models help to understand possible mechanisms behind variable timing of invasion by the Asian/American strain. The fitness gain resulting from the Asian/American strains shorter latency period increases as the adult vector mortality rate increases, indicating that regions where adult mosquito death rate is high will select for the more virulent strain of dengue 2.

Power consumption is an important constraint on VLSI systems. With the advancement in technology, it is now possible to pack a large range of functionalities into VLSI devices. Hence it is important to find out ways to utilize these functionalities with optimized power consumption. This work focuses on curbing power consumption at the design stage. This work emphasizes minimizing active power consumption by minimizing the load capacitance of the chip. Capacitance of wires and vias can be minimized using Ant Colony Optimization (ACO) algorithms. ACO provides a multi agent framework for combinatorial optimization problems and hence is used to handle multiple constraints of minimizing wire-length and vias to achieve the goal of minimizing capacitance and hence power consumption. The ACO developed here is able to achieve an 8% reduction of wire-length and 7% reduction in vias thereby providing a 7% reduction in total capacitance, compared to other state of the art routers.

Condition-specific cellular networks are networks of genes and proteins that describe functional interactions among genes occurring under different environmental conditions. These networks provide a systems-level view of how the parts-list (genes and proteins) interact within the cell as it functions under changing environmental conditions and can provide insight into mechanisms of stress response, cellular differentiation and disease susceptibility. The principle challenge, however, is that cellular networks remain unknown for most conditions and must be inferred from activity levels of genes (mRNA levels) under different conditions. This dissertation aims to develop computational approaches for inferring, analyzing and validating cellular networks of genes from expression data. This dissertation first describes an unsupervised machine learning framework for inferring cellular networks using expression data from a single condition. Here cellular networks are represented as undirected probabilistic graphical models and are learned using a novel, data-driven algorithm. Then several approaches are described that can learn networks using data from multiple conditions. These approaches apply to cases where the condition may or may not be known and, therefore, must be inferred as part of the learning problem. For the latter, the condition variable is allowed to influence expression of genes at different levels of granularity: condition variable per gene to a single condition variable for all genes. Results on simulated data suggest that the algorithm performance depends greatly on the size and number of connected components of the union network of all conditions. These algorithms are also applied to microarray data from two yeast populations, quiescent and non-quiescent, isolated from glucose starved cultures. Our results suggest that by sharing information across multiple conditions, better networks can be learned for both conditions, with many more biologically meaning

Zintl anions are best described as polyatomic main group clusters (i.e. E73-; E = P, As, Sb) with structures characteristic of isoelectronic clusters (i.e. hydrocarbon and borohydride clusters). Combining these main group clusters and transition-metal precursors (Ni(COD)2; COD = cyclooctadiene, and Pd(PCy3)2; PCy3 = tricyclohexyl phosphine) with very labile ligands allows for the isolation of novel binary molecular clusters. The freedom these structures experience due to the absence of organic ligands lends itself to the very unique structure growth the binary clusters exhibit. Direct oxidation of these binary molecular clusters results in the formation of binary phases (i.e. PdAs2, NiAs, NiAs2). More control of the phase-specific binaries obtained can be achieved by controlling the reaction of the precursors, eliminating the need to isolate the clusters. Binary nanoparticles are in the forefront of heterogeneous catalyst development and application, making this an area of intense investigation. Structural characteristics of the binary clusters, [As@Ni12@As20]3-, [(Ni2Sb2)(Sb7)2]4-, [Ni5Sb17]4-, [Pd7As16]4- and [Pd2(E7)2]4- (E = P, As), and evidence for the formation of phase-specific binary nanoparticles (Ni5As2, Ni11As8, NiAs and NiAs2) under very mild conditions will be presented.