Browsing by Subject "Network"

Integration of disparate research fields has become a major concern in recent years due to the increasing complexity of the issues that face policy makers and researchers. Concerted efforts have therefore been initiated to remove the traditional barriers between research fields to allow for greater cooperation between policy makers and researchers, particularly in the fields of health, the environment, and development. The Bridge Collaborative is one such organization dedicated to facilitating this process through the use of results chains. However, because of a lack of experimental data or observational datasets traditionally endemic to interdisciplinary policy research, they lack an effective mechanism for analyzing causal dependence among network variables. The purpose of this thesis is therefore to create a method of analyzing causal relationships using expert knowledge that can still pass the rigorous tests necessary to assert causality in the traditional experimental and observational data approaches. Building upon previous work of statisticians, philosophers, and computer scientists, I create a question template that will allow a researcher to easily check and refine a causal network and explore alternatives to that network based on experience and elicited expert judgement alone. I then perform a case study using this template based on the work of the Food-Energy-Water (FEW) Catalyst project, a group initiative within the Bridge Collaborative, to review a causal network based on a systematic literature search. I conclude that a causal network can indeed be constructed, explored, and adjusted using logical reasoning and expert judgement—a finding that has implications for researchers seeking to create reliable models using causal networks as their base.

This dissertation looks at how to identify and promote cooperation in a multiagent system, first theoretically through the lens of computational game theory and later empirically through a human subject experiment. Chapter 2 studies the network dynamics leading to a potential unraveling of cooperation and identify the subset of agents that can form an enforceable cooperative agreement with. This is an important problem, because cooperation is harder to sustain when information of defection, and thus the consequent punishment, transfers slowly through the network structures from a larger community. Chapter 3 examines a model that studies cooperation in a broader strategic context where agents may interact in multiple different domains, or games, simultaneously. Even if a game independently does not give an agent sufficient incentive to play the cooperative action, there may be hope for cooperation when multiple games with compensating asymmetries are put together. Exploiting compensating asymmetries, we can find an institutional arrangement that would either ensure maximum incentives for cooperation or require minimum subsidy to establish sufficient incentives for cooperation. Lastly, Chapter 4 studies a two-layered public good game to empirically examine whether community enforcement through existing bilateral relationships can encourage cooperation in a social dilemma situation. Here, it is found that how the situation is presented matters greatly to real life agents, as their understanding of whether they are in a cooperative or a competitive, strategic setting changes the level of overall cooperation.

Anticipation is a state of expectancy for something that will happen, and it allows us to use past learning to prepare for and make predictions about the future. Studies have shown that anticipation influences behavioral performance, learning, and memory, and studies implicate reward-related brain circuitry. However, few studies have investigated the neural underpinnings of anticipation on a brain-wide network scale . In this set of experiments, I take an interdisciplinary cross-species approach, using in-vivo electrophysiology in mice and functional magnetic resonance imaging (fMRI) in humans, to investigate brain networks underlying anticipation in motivated behavior. Using a data-driven machine learning approach, I characterize the anticipatory network in mice running through a T-maze, and show how it is affected by behavioral perturbation in the form of a task reversal, and circuit perturbation in the form of a genetic mutant mouse line. I also validate this network in a separate cohort of mice in a variation of the T-maze task that varies in difficulty, and show how activity in this network is modulated by task difficulty and intermediate instrumental goals. Finally, I investigate this network using fMRI in human subjects performing a trivia-based task to show how this network links curiosity, a more intrinsic form of motivation, to memory. The findings from these studies provide evidence at multiple levels and across multiple species for an anticipatory network that links motivational state to cognitive performance.

Accurate IP geo-location is crucial to the effectiveness of a wide array of Internet-based services ranging from targeted advertising and website localization to content delivery, security logging and authentication. The most widely used technique for remote IP geo-location is to passively query a pre-built database mapping IP blocks to physical locations. Recent analysis of commercially available databases has revealed limited global coverage and limited accuracy below the country level.

In this work, we first present a new form of geo-location technique which cross-references Regional Internet Registry (RIR) entries with topology information derived from Border Gateway Protocol (BGP) routing data. Second, we present a Hadoop integrated PATRICIA tree designed to store this dataset. Finally, we present a system for accurately and efficiently mapping location strings to representative alpha-shape polygons.

Our experiments show that cross-referencing RIR entries with topology information allows for improvements in location accuracy below the country level in comparison with traditional databases. Furthermore, we show that use of a PATRICIA tree provides maximum storage efficiency with minimal performance impact. Finally, we show that representing locations as alpha shapes provides a high level of accuracy with minimal performance overhead.

During embryogenesis, a single zygote gives rise to a multicellular embryo with distinct spatial territories marked by differential gene expression. How is this patterning process organized? How robust is this function to perturbations? Experiments that examine normal and regulative development will provide direct evidence for reasoning out the answers to these fundamental questions. Recent advances in technology have led to experimental determinations of increasingly complex gene regulatory networks (GRNs) underlying embryonic development. These GRNs offer a window into systems level properties of the developmental process, but at the same time present the challenge of characterizing their behavior. A suitable modeling framework for developmental systems is needed to help gain insights into embryonic development. Such models should contain enough detail to capture features of interest to developmental biologists, while staying simple enough to be computationally tractable and amenable to conceptual analysis. Combining experiments with the complementary modeling framework, we can grasp a systems level understanding of the regulatory program not readily visible by focusing on individual genes or pathways.

This dissertation addresses both modeling and experimental challenges. First, we present the autonomous Boolean network modeling framework and show that it is a suitable approach for developmental regulatory systems. We show that important timing information associated with the regulatory interactions can be faithfully represented in autonomous Boolean models in which binary variables representing expression levels are updated in continuous time, and that such models can provide direct insight into features that are difficult to extract from ordinary differential equation (ODE) models. As an application, we model the experimentally well-studied network controlling fly body segmentation. The Boolean model successfully generates the patterns formed in normal and genetically perturbed fly embryos, permits the derivation of constraints on the time delay parameters, clarifies the logic associated with different ODE parameter sets, and provides a platform for studying connectivity and robustness in parameter space. By elucidating the role of regulatory time delays in pattern formation, the results suggest new types of experimental measurements in early embryonic development. We then use this framework to model the much more complicated sea urchin endomesoderm specification system and describe our recent progress on this long term effort.

Second, we present experimental results on developmental plasticity of the sea urchin embryo. The sea urchin embryo has the remarkable ability to replace surgically removed tissues by reprogramming the presumptive fate of remaining tissues, a process known as transfating, which in turn is a form of regulative development. We show that regulative development requires cellular competence, and that competence is lost early on but can be regained after further differentiation. We demonstrate that regulative replacement of missing tissues can induce distal germ layers to participate in reprogramming, leading to a complete re-patterning in the remainder of the embryo. To understand the molecular mechanism of cell fate reprogramming, we examined micromere depletion induced non-skeletogenic mesoderm (NSM) transfating. We found that the skeletogenic program was greatly temporally compressed in this case, and that akin to another NSM transfating case, the transfating cells went through a hybrid regulatory state where NSM and skeletogenic marker genes were co-expressed.

Extensive investigation has been conducted on network data, especially weighted network in the form of symmetric matrices with discrete count entries. Motivated by statistical inference on multi-view weighted network structure, this paper proposes a Poisson-Gamma latent factor model, not only separating view-shared and view-specific spaces but also achieving reduced dimensionality. A multiplicative gamma process shrinkage prior is implemented to avoid over parameterization and efficient full conditional conjugate posterior for Gibbs sampling is accomplished. By the accommodating of view-shared and view-specific parameters, flexible adaptability is provided according to the extents of similarity across view-specific space. Accuracy and efficiency are tested by simulated experiment. An application on real soccer network data is also proposed to illustrate the model.

In recent years, the pervasiveness and importance of post-transcriptional regulation has reshaped the underlying principles of the organizational logic of gene expression. RNA-binding proteins (RBPs) and non-coding RNAs are the regulatory molecules primary responsible for interaction with target mRNAs and thereby regulating post-transcriptional processes eventually influencing characteristics of the encoded protein. Many of the mRNA targets of RBPs encode functionally related proteins, which for post-trascriptional operons, resulting in coordination of macromolecular complexes or specific cellular processes. Thus, identifying RNA targets, precise binding sites, and the dynamics of these interactions will reveal how these important regulatory factors contribute to gene regulatory networks.

ELAV family of human RBPs consist of 4 members, which all have 3 RRM (RNA-recognition motif) domains the last separated by a hinge region. It predominant role is to positively regulate the stability and translation of target mRNAs through binding to ARE (AU-rich elements) in the 3' UTR (untranslated region) of protein coding transcripts. In response to certain stimuli, HuR is subject to post-translational modifications and changes subcellular localization, which impacts its regulatory capacity. In this study on a transcriptome-wide level, we interrogate the RNA targets, precise binding sites, as well as the remodeling of these interactions in response to stimuli.

We utilized two complementary methods, RIP-chip and PAR-CLIP, to identify targets of HuR and high-resolution binding sites on a transcriptome-wide scale. We discovered that HuR-mRNA interactions are not restricted to the 3' UTR and there are thousands of intronic binding sites. A significant proportion of intronic binding sites are contained in the poly-pyrimidine tract near 3' splice sites. Binding sites in the 3' UTR and intron are often approximately 30 nucleotides apart. HuR can bind to both AU-rich and U-rich sequences, the former more prevalent in 3' UTRs and the latter more prevalent at the 3' splice site.

Next we integrated the binding data with transcriptomics of HuR siRNA mediated knockdown. We found that the degree of binding is proportional to the degree of HuR-dependent stabilization. Moreover the ability to stabilize mRNA is not restricted to 3' UTR binding sites, as intronic binding sites also exhibited the binding degree correlated stabilization. We observed that the spatial pattern of HuR binding sites relative to exons influences exon usage decisions. Specifically, binding sites upstream of the exon promote exclusion, while binding sites downstream of the exon promote inclusion.

Hybrid polymeric networks constructed using both covalent and reversible cross-links have been shown to be effective in preventing fracture and ultimately failure in polymeric materials. The prevention of failure has been largely attributed to the ability of the reversible cross-links to dissipate energy without breaking the covalent cross-links. The ability to rationally design materials that optimize this strategy would benefit from quantitative and systematic studies of the relationship between the number and strength of reversible interactions and the failure behavior of hybrid networks. This dissertation describe studies of fracture under compression in a family of hybrid networks, in which the timescale of reversible cross-linker dissociation is varied over several orders of magnitude, whereas the covalent components are kept constant.

Polymeric networks were constructed with 4-vinylpyridine. Bimetallic pincer Pd and Pt complexes were inserted into the network, forming reversible metal-ligand bonds that cross-link pyridine residues. The additional reversible cross-links prolong the lifetime of the hybrid networks under compressive strain when compared to their covalent counterparts. The observed failure behavior is dependent on the rate at which the networks are compressed as well as the strength of reversible interaction. Most interestingly, the addition of very dynamic and weak reversible interactions, so weak as to make no measurable contribution to bulk modulus, still leads to enhanced fracture strains. The failure of the covalent component within these hybrid networks was probed directly by incorporating a mechanophore that emits light upon chain scission. It was confirmed that the addition of these dynamic and weak reversible cross-links delays the catastrophic bond scission events associated with failure in the materials.

The Gram-negative bacteria Chlamydia trachomatis is a common sexually transmitted pathogen that can cause severe sequelae including cause pelvic inflammatory disease and sterility. This obligate intracellular pathogen effectively manipulates host cellular functions by secreting virulence factors across its membrane bound vacuole. Identifying these virulence components and how they help in establishing an environment conducive for bacterial growth is central to understanding chlamydial pathogenesis. This is experimentally challenging due to a lack of tools to perform molecular genetic studies. In the absence of genetic tools, we developed a yeast model system to identify and characterize chlamydial proteins involved in virulence mechanisms. In this study we describe the identification of twenty-eight proteins potentially involved in modulating host cellular functions and the secretion of virulence factors into the host. Since the delivery of virulence proteins by a type three secretion (T3S) system is a critical step for Chlamydia, we identified the proteins that interacted with the T3S apparatus by yeast two-hybrid analysis. We discovered several novel interactions between and determined that the C. trachomatis T3S apparatus displayed a similar architecture to that of other T3S systems. Furthermore with these approaches we identified networks of proteins that interacted with the secretion apparatus including a novel secretion chaperone protein. We characterized Ct260/Mcsc one of the putative secretion and demonstrated that it represents a novel class 1B secretion chaperone protein. Unlike other known chaperones, Mcsc directly interact with a conserved component of the T3S apparatus cytoplasmic domain, CdsQ. These finding represents a novel mechanism by which the secretion chaperone protein Ct260 may increase the secretion efficiency of its effector cargo and may reveal new facets of secretory cargo recognition by T3S systems.

With the advent of Web 2.0 and the Digital Age, we are witnessing an unprecedented increase in the amount of information collected, and in the number of users interested in different types of information. This growth means that traditional techniques, where users poll data sources for information of interest, are no longer sufficient. Polling too frequently does not scale, while polling less often may result in users missing important updates. The alternative push technology has long been the goal of publish/subscribe systems, which proactively push updates (events) to users with matching interests (expressed as subscriptions). The push model is better suited for ensuring scalability and timely delivery of updates, important in many application domains: personal (e.g., RSS feeds, online auctions), financial (e.g., portfolio monitoring), security (e.g., reporting network anomalies), etc.

Early publish/subscribe systems were based on predefined subjects (channels), and were too coarse-grained to meet the specific interests of different subscribers. The second generation of content-based publish/subscribe systems offer greater flexibility by supporting subscriptions defined as predicates over message contents. However, subscriptions are still stateless filters over individual messages, so they cannot express queries across different messages or over the event history. The few systems that support more powerful database-style subscriptions do not address the problem of efficiently delivering updates to a large number of subscribers over a wide-area network. Thus, there is a need to develop next-generation publish/subscribe systems that unify the support for richer database-style subscription queries and flexible wide-area notification. This support needs to be complemented with robust processing and dissemination techniques that scale to high event rates and large databases, as well as to a large number of subscribers over the Internet.

The main contribution of our work is a collection of techniques to support efficient and scalable event processing and notification dissemination for an Internet-scale publish/subscribe system with a rich subscription model. We investigate the interface between event processing by a database server and notification delivery by a dissemination network. Previous research in publish/subscribe has largely been compartmentalized; database-centric and network-centric approaches each have their own limitations, and simply putting them together does not lead to an efficient solution. A closer examination of database/network interfaces yields a spectrum of new and interesting possibilities. In particular, we propose message and subscription reformulation as general techniques to support stateful subscriptions over existing content-driven networks, by converting them into equivalent but stateless forms. We show how reformulation can successfully be applied to various stateful subscriptions including range-aggregation, select-joins, and subscriptions with value-based notification conditions. These techniques often provide orders-of-magnitude improvement over simpler techniques adopted by current systems, and are shown to scale to millions of subscriptions. Further, the use of a standard off-the-shelf content-driven dissemination interface allows these techniques to be easily deployed, managed, and maintained in a large-scale system.

Based on our findings, we have built a high-performance publish/subscribe system named ProSem (to signify the inseparability of database processing and network dissemination). ProSem uses our novel techniques for group-processing many types of complex and expressive subscriptions, with a per-event optimization framework that chooses the best processing and dissemination strategy at runtime based on online statistics and system objectives.