Luminescent molecules and polymers are active component in light-emitting diodes; however, the aggregation and excimer formation in concentrated solution or in the solid film states had limited their applications. Therefore, this study used poly(methyl methacrylate) (PMMA) as separator to prevent the formations of aggregate and excimer and to enhance quantum efficiency. Basically, two systems are involved: (1) Inorganic phosphorescent irdium complex PMMA was doped with inorganic iridium complex IrQB by using THF as solvent. IrQB/PMMA films prepared from dilute solutions exhibit two emissions centered at 560 and 640 nm, respectively; in contrast, only 640-nm emission was observed for films from concentrated solutions. Experimentally, these two bands showed variations on the emission intensity with increasing temperature. Aggregation of IrQB is suggested to be responsible for the 560-nm emission. Chain conformation of PMMA in the solution state strongly affects the incorporated IrQBs and their emission properties. (2) Polybenzoxazines Polybenzoxazines with the built-in fluorenscent fluorine moiety are linear in nature; however, the inherent hydrogen-bond (H-bond) interactions in polybenzoxazines decrease the inter-chain distance and cause the chain aggregation. With the added PMMAs, new H-bonding from the carbonyl groups in PMMA and the hydroxyl groups in polybenzoxazine enhances the mutual miscibility between these two components and decreases the possibility of aggregate formation in polybenzoxazines. Quantum efficiency is therefore promoted by this approach.

Platelets play a pivotal role in hemostasis and thrombosis. It induces vascular retraction and clot formation through platelets activation and signal transmission, which promote ligands expressing on the surface of platelets, such as glycoproteins and P-selectin. Some surface glycoproteins gatherd to form complexes after activation. It bound to extracellular receptors such as collagen and thrombin to induce aggregation, which could also be induced by releasing agonists, such as arachidonic acid, adenosine diphosphate, epinephrine, serotonin and fibrinogen, to active the nearby platelets. The standard process of plateletapheresis in the Blood Center was to hold the platelets in still for one hour before stored on a vibrator. The process of holding platelets still for one hour before storage was omitted in some hospitals. It was not clear whether to omit the process has any effect on the quality of platelets. The expressions of P-selectin and vWF receptor, CD42b (Gp Ib£\) on platelets were analyzed by flow cytometry in this study. No significant differences (p= 0.77 and p= 0.62, respectively) were found. Similar results were obtained when functions of platelets were evaluated by agonists. It was concluded that leaving the platelets in room temperature for one hour to recover before keeping it on a vibrator would not enhance the functions of platelets aggregation significantly.

Amyloidoses comprise a group of diseases where normal or mutated protein precipitates into amyloid fibrils. The deposition of fibrils causes dysfunction of organs and toxicity to nervous tissue. Up to date, 24 different proteins and peptides are known to be able to form amyloid fibrils. The most well known are Amyloid beta peptide and Prione protein causing Alzheimer’s disease and Creutzfeld Jacob’s disease respectively. The aims of this thesis were to investigate the structural properties of cytotoxic amyloid and examine the mechanisms involved. The model protein mostly used in the studies was the plasma protein transthyretin (TTR). Familial Amyloidotic Polyneuropathy (FAP) is a hereditary, autosomal-dominant neurodegenerative disease caused by point mutations in the TTR gene. One of the most common variants of FAP is a mutation in position 30 where alanine is exchanged for methonine. This gives rise to “Skellefteåsjukan” in Sweden. TTR is secreted into the plasma as a tetramer. Point mutations destabilize the tetramer leading to disassembled monomers, which undergo partial denaturation as an initiation step to aggregation and amyloid fibril formation. In vivo amyloidogenesis takes a long time and does not occur until late in adult life. Most of the clinical TTR mutations do not form amyloid in vitro under physiological conditions. We have created amyloidogenic TTR mutants that are prone to aggregate and form fibrils under physiological conditions. This provides us with a model system on the cellular level for studies of the mechanisms of amyloid associated cytotoxicity as we can control the aggregation process and capture defined stages in the TTR amyloidogenic pathway. We used Atomic Force Microscopy (AFM) to follow the morphology of aggregates during fibril formation. Initially, amorphous aggregates were formed that subsequently matured into fibrillar structures, denoted protofilaments. This observation was interpreted as an optimisation of ß-strand registers. In addition we identified a correlation between the presence of early-formed aggregates of TTR and cytotoxicity. The toxic response was mediated via an apoptotic mechanism. We were not able to more carefully determine the structure and size of the toxic TTR species. To address this problem we turned to another amyloidogenic protein, equine lysozyme (EL). Intermediate samples corresponding to the aggregation and growth phase of amyloid fibrils of EL were collected. These samples were subjected to cytotoxicity assays as well as monomeric starting material and mature amyloid fibrillar species. The results clearly showed that the soluble oligomers were cytotoxic in contrast to the monomers and fibrils. Our data indicate that the toxic properties of the oligomers are size dependent. In this thesis we asked the question whether all mutated forms of TTR can be expressed and secreted or if there is a selection against the most aggressive mutations in vivo? We transfected hematopoetic K562 cells with wild type or mutant TTR, with or without the N-terminal signal peptide, responsible for secretion, to generate both extra- and intracellular TTR. We show that the post-translational quality control of the cells does not allow intracellular mutant TTR outside the secretory pathway, possibly due to the cytotoxic effects, while translocated to the secretory pathway made it escape the quality control permitting secretion and amyloid formation outside the cells. We have further analyzed the cytotoxic mechanisms induced by TTR oligomers with a focus on intracellular apoptotic signalling pathways. We show that TTR oligomers bind to the surface of the target cells but are not taken up, that is in contrast to mature fibrils that do not bind them at all. The apoptotic response occurred in a caspase-independent and a free radical dependent way.

Most existing data mining approaches perform data mining tasks on a single data table. However, increasingly, data repositories such as financial data and medical records, amongst others, are stored in relational databases. The inability of applying traditional data mining techniques directly on such relational database thus poses a serious challenge. To address this issue, a number of researchers convert a relational database into one or more flat files and then apply traditional data mining algorithms. The above-mentioned process of transforming a relational database into one or more flat files usually involves aggregation. Aggregation functions such as maximum, minimum, average, standard deviation, count and sum are commonly used in such a flattening process. Our research aims to address the following question: Is there a link between aggregation and possible privacy violations during relational database mining? In this research we investigate how, and if, applying aggregation functions will affect the privacy of a relational database, during supervised learning, or classification, where the target concept is known. To this end, we introduce the PBIRD (Privacy Breach Investigation in Relational Databases) methodology. The PBIRD methodology combines multi-view learning with feature selection, to discover the potentially dangerous sets of features as hidden within a database. Our approach creates a number of views, which consist of subsets of the data, with and without aggregation. Then, by identifying and investigating the set of selected features in each view, potential privacy breaches are detected. In this way, our PBIRD algorithm is able to discover those features that are correlated with the classification target that may also lead to revealing of sensitive information in the database. Our experimental results show that aggregation functions do, indeed, change the correlation between attributes and the classification target. We show that with aggregation, we obtain a set of features which can be accurately linked to the classification target and used to predict (with high accuracy) the confidential information. On the other hand, the results show that, without aggregation we obtain another different set of potentially harmful features. By identifying the complete set of potentially dangerous attributes, the PBIRD methodology provides a solution where the database designers/owners can be warned, to subsequently perform necessary adjustments to protect the privacy of the relational database. In our research, we also perform a comparative study to investigate the impact of aggregation on the classification accuracy and on the time required to build the models. Our results suggest that in the case where a database consists only of categorical data, aggregation should especially be used with caution. This is due to the fact that aggregation causes a decrease in overall accuracies of the resulting models. When the database contains mixed attributes, the results show that the accuracies without aggregation and with aggregation are comparable. However, even in such scenarios, schemas without aggregation tend to slightly outperform. With regard to the impact of aggregation on the model building time, the results show that, in general, the models constructed with aggregation require shorter building time. However, when the database is small and consists of nominal attributes with high cardinality, aggregation causes a slower model building time.

This thesis focuses on three fundamental issues that concern data aggregation protocols for periodic data collection in sensor networks: <i>which</i> sensor nodes should report their data, <i>when</i> should they report it, and should they use <i>unicast</i> or <i>broadcast</i> based protocols for this purpose. <p> The issue of when nodes should report their data is considered in the context of real-time monitoring applications. The first part of this thesis shows that asynchronous aggregation, in which the time of each nodes transmission is determined adaptively based on its local history of past packet receptions from its children, outperforms synchronous aggregation by providing lower delay for a given end-to-end loss rate. <p> Second, new broadcast-based aggregation protocols that minimize the number of packet transmissions, relying on multipath delivery rather than automatic repeat request for reliability, are designed and evaluated. The performance of broadcast-based aggregation is compared to that of unicast-based aggregation, in the context of both real-time and delay-tolerant data collection. <p> Finally, this thesis investigates the potential benefits of dynamically, rather than semi-statically, determining the set of nodes reporting their data, in the context of applications in which coverage of some monitored region is to be maintained. Unicast and broadcast-based coverage-preserving data aggregation protocols are designed and evaluated. The performance of the proposed protocols is compared to that of data collection protocols relying on node scheduling.

scyllo-Inositol is currently in phase II clinical trials as a therapeutic for Alzheimer’s disease (AD). Previous work from our lab has shown that scyllo-inositol prevents Ab1-42 fibril formation instead leading to the formation of small Ab oligomers in vitro. To further understand the molecular details of Ab-scyllo-inositol binding interactions, a library of scyllo-inositol derivatives was prepared. A sequence of protecting group transformations afforded a hydroxylamine functionalized scyllo-inositol. Subsequent oxime formation with aromatic aldehydes generated a novel class of inositol derivatives in good yield and high purity. The effects of these compounds on the Ab aggregation cascade were evaluated by a biotin-avidin Ab1-42 oligomer assay and atomic force microscopy (AFM). Preliminary plate assay data indicated that several of these derivatives increased peptide oligomerization and the corresponding AFM images showed altered fibril formation. These results suggested that this class of scyllo-inositol derivatives is active in the Ab aggregation cascade.

scyllo-Inositol is currently in phase II clinical trials as a therapeutic for Alzheimer’s disease (AD). Previous work from our lab has shown that scyllo-inositol prevents Ab1-42 fibril formation instead leading to the formation of small Ab oligomers in vitro. To further understand the molecular details of Ab-scyllo-inositol binding interactions, a library of scyllo-inositol derivatives was prepared. A sequence of protecting group transformations afforded a hydroxylamine functionalized scyllo-inositol. Subsequent oxime formation with aromatic aldehydes generated a novel class of inositol derivatives in good yield and high purity. The effects of these compounds on the Ab aggregation cascade were evaluated by a biotin-avidin Ab1-42 oligomer assay and atomic force microscopy (AFM). Preliminary plate assay data indicated that several of these derivatives increased peptide oligomerization and the corresponding AFM images showed altered fibril formation. These results suggested that this class of scyllo-inositol derivatives is active in the Ab aggregation cascade.

Investigation of the NIR cyanine dye MHI-36 shows binding affinity to charged amino acids. This cyanine dye showed aggregation and dimer formation at higher dye concentration (2.0x10-3 M) induced by lysine. When dye concentration decreased to 1.0x10-4M no strong aggregate formation was viewed. Dye shows strong binding and selectivity properties towards charged amino acids lysine and arginine, compared to neutral leucine. It’s believed the positively charged presence was able to break and disrupt the conjugated π- π bonds at lower dye concentration. Computational work showed intramolecular aggregation of the phenyl groups on the dye. These aggregates are believed to create electron rich environment suitable for lysine interaction.

Most existing data mining approaches perform data mining tasks on a single data table. However, increasingly, data repositories such as financial data and medical records, amongst others, are stored in relational databases. The inability of applying traditional data mining techniques directly on such relational database thus poses a serious challenge. To address this issue, a number of researchers convert a relational database into one or more flat files and then apply traditional data mining algorithms. The above-mentioned process of transforming a relational database into one or more flat files usually involves aggregation. Aggregation functions such as maximum, minimum, average, standard deviation, count and sum are commonly used in such a flattening process. Our research aims to address the following question: Is there a link between aggregation and possible privacy violations during relational database mining? In this research we investigate how, and if, applying aggregation functions will affect the privacy of a relational database, during supervised learning, or classification, where the target concept is known. To this end, we introduce the PBIRD (Privacy Breach Investigation in Relational Databases) methodology. The PBIRD methodology combines multi-view learning with feature selection, to discover the potentially dangerous sets of features as hidden within a database. Our approach creates a number of views, which consist of subsets of the data, with and without aggregation. Then, by identifying and investigating the set of selected features in each view, potential privacy breaches are detected. In this way, our PBIRD algorithm is able to discover those features that are correlated with the classification target that may also lead to revealing of sensitive information in the database. Our experimental results show that aggregation functions do, indeed, change the correlation between attributes and the classification target. We show that with aggregation, we obtain a set of features which can be accurately linked to the classification target and used to predict (with high accuracy) the confidential information. On the other hand, the results show that, without aggregation we obtain another different set of potentially harmful features. By identifying the complete set of potentially dangerous attributes, the PBIRD methodology provides a solution where the database designers/owners can be warned, to subsequently perform necessary adjustments to protect the privacy of the relational database. In our research, we also perform a comparative study to investigate the impact of aggregation on the classification accuracy and on the time required to build the models. Our results suggest that in the case where a database consists only of categorical data, aggregation should especially be used with caution. This is due to the fact that aggregation causes a decrease in overall accuracies of the resulting models. When the database contains mixed attributes, the results show that the accuracies without aggregation and with aggregation are comparable. However, even in such scenarios, schemas without aggregation tend to slightly outperform. With regard to the impact of aggregation on the model building time, the results show that, in general, the models constructed with aggregation require shorter building time. However, when the database is small and consists of nominal attributes with high cardinality, aggregation causes a slower model building time.

Aggregation of multiple Reinforcement Learning (RL) algorithms is a new and effective technique to improve the quality of Sequential Decision Making (SDM). The quality of a SDM depends on long-term rewards rather than the instant rewards. RL methods are often adopted to deal with SDM problems. Although many RL algorithms have been developed, none is consistently better than the others. In addition, the parameters of RL algorithms significantly influence learning performances. There is no universal rule to guide the choice of algorithms and the setting of parameters. To handle this difficulty, a new multiple RL system - Aggregated Multiple Reinforcement Learning System (AMRLS) is developed. In AMRLS, each RL algorithm (learner) learns individually in a learning module and provides its output to an intelligent aggregation module. The aggregation module dynamically aggregates these outputs and provides a final decision. Then, all learners take the action and update their policies individually. The two processes are performed alternatively. AMRLS can deal with dynamic learning problems without the need to search for the optimal learning algorithm or the optimal values of learning parameters. It is claimed that several complementary learning algorithms can be integrated in AMRLS to improve the learning performance in terms of success rate, robustness, confidence, redundance, and complementariness. There are two strategies for learning an optimal policy with RL methods. One is based on Value Function Learning (VFL), which learns an optimal policy expressed as a value function. The Temporal Difference RL (TDRL) methods are examples of this strategy. The other is based on Direct Policy Search (DPS), which directly searches for the optimal policy in the potential policy space. The Genetic Algorithms (GAs)-based RL (GARL) are instances of this strategy. A hybrid learning architecture of GARL and TDRL, HGATDRL, is proposed to combine them together to improve the learning ability. AMRLS and HGATDRL are tested on several SDM problems, including the maze world problem, pursuit domain problem, cart-pole balancing system, mountain car problem, and flight control system. Experimental results show that the proposed framework and method can enhance the learning ability and improve learning performance of a multiple RL system.