The task of medicinal chemists in a drug discoveryproject is to synthesize/design analogues to the screening hits, simultaneouslyincreasing target potency and optimizing the pharmacological properties.  This requires a wide selection of moleculesto be synthesized, where both synthetic feasibility and price of startingmaterials are of great importance. In this work, a synthetic pathway from cheapand readily available starting materials to highly modifiable 2,4-disubstitutedpyrrolidines is demonstrated. Previously reported procedures to similarpyrrolidines use expensive catalysts, requires harsh conditions and requiresnon-commercially available starting materials. The suggested pathway herein has demonstrated great possibility forvariation in the 4-position, including fluoro, difluoro, nitrile and alcoholfunctional groups. There are several areas in which the synthesis can beimproved and expanded upon. Improvements can be made by optimizing thedescribed reaction conditions and further expansion of possible modificationsin both 2- and 4-position could be explored.

It is a challenging task for interaction designers to find a way to design a digital artefact supporting serendipitous discovery. Its interdisciplinary nature requires sufficient knowledge of information visualization, social navigation and serendipity. Based on literature review and prior relevant works, several traces having potential to aid such exploration were defined. Through creating and testing AniMap, an interactive graph visualization for discovering new anime clips, in this thesis I argue that such an artefact has the potential to support serendipitous discovery, owing to its features of being information visualization, interactive and in a graph layout, coupled with users’ personal interests. Even so, finding details of how to influence serendipitous discovery remain an ongoing challenge considering the dynamic nature of serendipity.

<p><strong>A growing body of evidence suggests that most human tumors, including those of the breast, are organized as cellular hierarchies. Positioned at the apex of these hierarchies are tumor-initiating cells (TICs), which are capable of limitless self-renewal and also differentiate, to give rise to various populations of non-tumorigenic cells that make up the bulk of the tumor. Importantly, recent findings have demonstrated that TICs are refractory to current best practice therapies, and thus likely account for high rates of tumor recurrence following remission. Therefore, it will likely be important to identify novel means of targeting TICs in order to achieve durable cancer cures.</strong></p> <p><strong>Using a highly sensitive transplantation assay, our laboratory previously showed that mammary tumors arising in various strains of transgenic mice comprise a very high fraction of TICs, and that when cells from these tumors are propagated in serum-free medium as tumorspheres, the high frequency of TICs is maintained. We therefore sought to use mouse mammary tumorspheres as an <em>in vitro</em> system with which to identify TIC-targeted agents and carried out a high-throughput screen of nearly 32,000 small molecules. To eliminate compounds showing general toxicity, we employed mouse mammospheres, which primarily comprise normal mammary epithelial stem and progenitor cells, in a secondary screen. Using this platform, we identified a small molecule that selectively targeted tumorsphere-derived cells <em>in vitro</em> and led to tumor growth arrest and tumor cell death <em>in vivo</em>. This study illustrates the utility of mouse models and high throughput screening to identify compounds which may target TICs but spare untransformed stem cells.</strong></p> / Master of Science (MSc)

<p>Nonribosomal peptides (NRPs) are a major class of natural products known for their biological activities and are employed therapeutically as immunosupressants, anticancer agents, and antibiotics. Nonribosomal peptides are microbial products, biosynthesized by large assembly line-like enzymes, known as nonribosomal peptide synthetases (NRPSs) that can be found in large gene clusters within the genome. With the advent of genome sequencing, the gene clusters for known NRPs are easily identified within producing organisms, but more strikingly, this sequencing reveals that microbes often contain many gene clusters with no known products suggesting traditional methods of isolation are overlooking the majority of NRPs.</p> <p>Extensive studies of NRPS functions have revealed assembly line logic for the biosynthesis of NRPs and using this knowledge, the NRP products of NRPS gene clusters can be predicted. In this research, products from both a simple dimodular NRPS from <em>Staphylococcus aureus </em>and a complex 11 module NRPS from <em>Delftia acidovorans </em>were predicted and used to successfully identify and isolate two novel NRPs, aureusimine and delftibactin.<em> </em>Theses compounds fell outside traditional NRP activities, one being a virulence regulator and the other a gold-complexing metallophore. Subsequent biosynthetic studies of the aureusimine gene cluster within the heterologous host, <em>Escherichia coli</em>, provide insight into NRPS flexibility for the creation of NRP natural variants and highlighted the utility of <em>E. coli </em>for the heterologous production of NRPs.</p> <p>Realizing single NRP predictions are not always accurate, a strategy was devised to use a genomically predicted NRP fragment barcode databases with the LC-MS/MS dereplication algorithm, iSNAP, to chemoinformatically identify and physically locate genetically predicted NRPs within crude extracts. This final contribution eliminates the need for bioactivity guided approaches to discovery and provides a strategy to systematically discover all predicted NRPs from cryptic gene clusters. This thesis delivers strategies and technologies for the directed discovery of NRPs from microbial sources.</p> / Doctor of Philosophy (PhD)

While numerous studies have investigated the impact of oil prices on the stock market, Chapter 2 is the first to examine the association between corporate bond yields and oil returns. We examine the association between oil-returns and corporate bond yields of four major U.S. industrial and financial sectors (including thirteen sub-sectors). Chapter 3 examines the reaction of stock markets in the U.K. and the Netherlands to a major composite event in the oil industry – the merger of the Royal Dutch Shell (RDSA) and the BG Group (BRGYY) on April 8, 2015, and the subsequent discovery of oil in southern England on April 9. We employ an exponential autoregressive conditionally heteroskedastic (EGARCH (1, 1)) framework in both Chapters, which allows for asymmetry of the effects between positive and negative external shocks including oil return shocks, shows the effects on both the yields/stock returns and their volatilities, and permits the persistence of the shocks to be measured. Three main results are obtained in Chapter 2. First, oil returns are significantly associated with the yield levels of corporate bonds issued in ten out of the thirteen sub-sectors considered within the oil-substitute, oil-related, oil-user, and financial services sectors. The three exceptions are the Petroleum Refinery, Building, and Chemical sub-sectors. Second, the return volatilities of corporate bonds issued in the Plastic & Rubber sub-sector demonstrate asymmetric responses to positive and negative shocks. To elaborate, negative shocks lead to lower volatility in the Plastic & Rubber sub-sector than positive shocks of the same magnitude. Third, the half-life, or the time it takes for the volatility of the portfolio of bonds in the Industrial Machinery sub-sector to move halfway back to its conditional mean after a shock is introduced, is 8.6 months. For bonds in all other sub-sectors, the half-life is less than 2.5 months. We obtain several results in Chapter 3. First, the composite event of merger and oil discovery generated significant abnormal returns in six out of the thirteen sub-sectors considered in the U.K. and three out of ten sub-sectors in the Netherlands. The remaining seven sub-sectors in the U.K. and the other seven sub-sectors in the Netherlands show no sensitivity in returns to the shock. Second, there is evidence of some information leakage about the composite event as demonstrated in the significant abnormal returns for Coal, Oil & Gas Extraction, Depository Institute, Chemical and Plastic & Rubber sub-sectors in U.K. and Coal, Depository Institute and Air Transportation sub-sectors in the Netherlands up to three days before the announcement of the composite event. Third, the behavioral patterns of four of the thirteen sub-sectors considered in the U.K. and four of the ten sub-sectors considered in the Netherlands demonstrate asymmetry in response to external shocks to their respective returns. These results have three main implications. First, investors holding bonds issued by the two sub-sectors with asymmetric oil shock effects need to add bonds from oil-related and oil-substitute sectors to lower the volatility of their bond portfolio because the latter do not exhibit asymmetry. Second, considering the overall finding of sensitivity to oil price changes, institutional investors need to examine the sensitivity of their bond portfolios to oil return changes and to guard against excessive risk. Similarly, corporations should monitor oil price variations and hedge the volatility risk accordingly. Finally, stock investors in the U.K. and the Netherlands might benefit from monitoring the key events that may affect the oil supply and oil prices, and acting accordingly. / Economics

The false discovery rate (FDR) introduced by Benjamini and Hochberg (1995) is perhaps the most standard error controlling measure being used in a wide variety of applications involving multiple hypothesis testing. There are two approaches to control the FDR - the fixed error rate approach of Benjamini and Hochberg (BH, 1995) where a rejection region is determined with the FDR below a fixed level and the estimation based approach of Storey (2002) where the FDR is estimated for a fixed rejection region before it is controlled. In this proposal, we concentrate on both these approaches and propose new, improved versions of some FDR controlling procedures available in the literature. A number of adaptive procedures have been put forward in the literature, each attempting to improve the method of Benjamini and Hochberg (1995), the BH method, by incorporating into this method an estimate of number true null hypotheses. Among these, the method of Benjamini, Krieger and Yekutieli (2006), the BKY method, has been receiving lots of attention recently. In this proposal, a variant of the BKY method is proposed by considering a different estimate of number true null hypotheses, which often outperforms the BKY method in terms of the FDR control and power. Storey's (2002) estimation based approach to controlling the FDR has been developed from a class of conservatively biased point estimates of the FDR under a mixture model for the underlying p-values and a fixed rejection threshold for each null hypothesis. An alternative class of point estimates of the FDR with uniformly smaller conservative bias is proposed under the same setup. Numerical evidence is provided to show that the mean squared error (MSE) is also often smaller for this new class of estimates. Compared to Storey's (2002), the present class provides a more powerful estimation based approach to controlling the FDR. / Statistics

The g - and - h distributional family is generated from a relatively simple transformation of the standard normal. By changing the skewness and elongation parameters g and h, this distributional family can approximate a broad spectrum of commonly used distributional shapes, such as normal, lognormal, Weibull and exponential. Consequently, it is easy to use in simulation studies and has been applied in multiple areas, including risk management, stock return analysis and missing data imputation studies. The current available methods to estimate the g - and - h distributional family include: letter value based method (LV), numerical maximal likelihood method (NMLE), and moment methods. Although these methods work well when no outliers or contaminations exist, they are not resistant to a moderate amount of contaminated observations or outliers. Meanwhile, NMLE is a computational time consuming method when data sample size is large. In this dissertation a quantile based least squares (QLS) estimation method is proposed to fit the g - and - h distributional family parameters and then derive its basic properties. Then QLS method is extended to a robust version (rQLS). Simulation studies are performed to compare the performance of QLS and rQLS methods with LV and NMLE methods to estimate the g - and - h parameters from random samples with or without outliers. In random samples without outliers, QLS and rQLS estimates are comparable to LV and NMLE in terms of bias and standard error. On the other hand, rQLS performs better than other non-robust method to estimate the g - and - h parameters when moderate amount of contaminated observations or outliers exist. The flexibility of the g - and - h distribution and the robustness of rQLS method make it a useful tool in various fields. The boxplot (BP) method had been used in multiple outlier detections by controlling the some-outside rate, which is the probability of one or more observations, in an outlier-free sample, falling into the outlier region. The BP method is distribution dependent. Usually the random sample is assumed normally distributed; however, this assumption may not be valid in many applications. The robustly estimated g - and - h distribution provides an alternative approach without distributional assumptions. Simulation studies indicate that the BP method based on robustly estimated g - and - h distribution identified reasonable number of true outliers while controlling number of false outliers and some-outside rate compared to normal distributional assumption when it is not valid. Another application of the robust g - and - h distribution is as an empirical null distribution in false discovery rate method (denoted as BH method thereafter). The performance of BH method depends on the accuracy of the null distribution. It has been found that theoretical null distributions were often not valid when simultaneously performing many thousands, even millions, of hypothesis tests. Therefore, an empirical null distribution approach is introduced that uses estimated distribution from the data. This is recommended as a substitute to the currently used empirical null methods of fitting a normal distribution or another member of the exponential family. Similar to BP outlier detection method, the robustly estimated g - and - h distribution can be used as empirical null distribution without any distributional assumptions. Several real data examples of microarray are used as illustrations. The QLS and rQLS methods are useful tools to estimate g - and - h parameters, especially rQLS because it noticeably reduces the effect of outliers on the estimates. The robustly estimated g - and - h distributions have multiple applications where distributional assumptions are required, such as boxplot outlier detection or BH methods. / Statistics

PhD Thesis / Metabolomics offers a systemic approach to discover clinical biomarkers for early detection of chronic diseases while also revealing underlying mechanisms relevant to human disorders of complex aetiology. Metabolomic studies in support of chronic disease prevention have focused primarily on surrogate biofluids (e.g., serum, plasma) for analysis due to their routine and less invasive sample collection in a clinical setting. However, biofluids are non-organ specific and thus are reflective of confounding biochemical processes within the body that are often difficult to interpret. As a result, it is necessary to assess metabolite changes localized within tissues since they are the direct site of pathogenic processes, in order to obtain more robust and specific biomarkers. This thesis aims to contribute to new advances in biomarker discovery and tissue metabolomic studies using multiplexed separations together with innovative data workflows based on multisegment injection-capillary electrophoresis-mass spectrometry (MSI-CE-MS). Chapter II introduces a high throughput yet targeted screening method for accurate quantification of serum γ‐glutamyl dipeptides from a cohort of overweight Japanese non-alcoholic steatohepatitis (NASH) patients that may allow for better risk assessment of long-term survivorship complementary to histopathology. Chapter III introduces a non-targeted metabolite profiling strategy for fasting plasma samples from prediabetic, older adults undergoing short-term step reduction (<1000 steps/day) in order to identify adaptive metabolic responses to abrupt changes in physical inactivity for early detection of sarcopenia in high-risk older persons. Chapter IV describes the first metabolomics study to characterize the human skeletal muscle metabolome from mass-restricted tissue biopsies together with matching plasma samples, which identified novel metabolic signatures associated with strenuous interval exercise, as well as treatment effects from high-dose bicarbonate pretreatment that delays the onset of muscle fatigue. Lastly, in Chapter V, metabolite coverage was expanded to include fatty acids for comprehensive characterization of murine placental tissue metabolome, which revealed sex-specific metabolic adaptations during gestation from maternal dams fed a standardized diet. In summary, this thesis contributes to new innovations in metabolomics for the discovery of novel biomarkers from blood and/or tissue specimens as required for early detection of chronic diseases relevant to population health, which were also used to validate the efficacy of therapeutic interventions based on physical activity to support healthy ageing. / Thesis / Doctor of Philosophy (PhD)

no / While criminals may start their activities at individual level, the same is in general not true for terrorists who are mostly organized in well established networks. The effectiveness of a terror network could be realized by watching many factors, including the volume of activities accomplished by its members, the capabilities of its members to hide, and the ability of the network to grow and to maintain its influence even after the loss of some members, even leaders. Social network analysis, data mining and machine learning techniques could play important role in measuring the effectiveness of a network in general and in particular a terror network in support of the work presented in this chapter. We present a framework that employs clustering, frequent pattern mining and some social network analysis measures to determine the effectiveness of a network. The clustering and frequent pattern mining techniques start with the adjacency matrix of the network. For clustering, we utilize entries in the table by considering each row as an object and each column as a feature. Thus features of a network member are his/her direct neighbors. We maintain the weight of links in case of weighted network links. For frequent pattern mining, we consider each row of the adjacency matrix as a transaction and each column as an item. Further, we map entries into a 0/1 scale such that every entry whose value is greater than zero is assigned the value one; entries keep the value zero otherwise. This way we can apply frequent pattern mining algorithms to determine the most influential members in a network as well as the effect of removing some members or even links between members of a network. We also investigate the effect of adding some links between members. The target is to study how the various members in the network change role as the network evolves. This is measured by applying some social network analysis measures on the network at each stage during the development. We report some interesting results related to two benchmark networks: the first is 9/11 and the second is Madrid bombing.

Systems biology comprises the global, integrated analysis of large-scale data encoding different levels of biological information with the aim to obtain global insight into the cellular networks. Several studies have unveiled the modular and hierarchical organization inherent in these networks. In this dissertation, we propose and develop innovative systems approaches to integrate multi-source biological data in a modular manner for network inference and biomarker discovery in complex diseases such as breast cancer. The first part of the dissertation is focused on gene module identification in gene expression data. As the most popular way to identify gene modules, many cluster algorithms have been applied to the gene expression data analysis. For the purpose of evaluating clustering algorithms from a biological point of view, we propose a figure of merit based on Kullback-Leibler divergence between cluster membership and known gene ontology attributes. Several benchmark expression-based gene clustering algorithms are compared using the proposed method with different parameter settings. Applications to diverse public time course gene expression data demonstrated that fuzzy c-means clustering is superior to other clustering methods with regard to the enrichment of clusters for biological functions. These results contribute to the evaluation of clustering outcomes and the estimations of optimal clustering partitions. The second part of the dissertation presents a hybrid computational intelligence method to infer gene regulatory modules. We explore the combined advantages of the nonlinear and dynamic properties of neural networks, and the global search capabilities of the hybrid genetic algorithm and particle swarm optimization method to infer network interactions at modular level. The proposed computational framework is tested in two biological processes: yeast cell cycle, and human Hela cancer cell cycle. The identified gene regulatory modules were evaluated using several validation strategies: 1) gene set enrichment analysis to evaluate the gene modules derived from clustering results; (2) binding site enrichment analysis to determine enrichment of the gene modules for the cognate binding sites of their predicted transcription factors; (3) comparison with previously reported results in the literatures to confirm the inferred regulations. The proposed framework could be beneficial to biologists for predicting the components of gene regulatory modules in which any candidate gene is involved. Such predictions can then be used to design a more streamlined experimental approach for biological validation. Understanding the dynamics of these gene regulatory modules will shed light on the related regulatory processes. Driven by the fact that complex diseases such as cancer are “diseases of pathways”, we extended the module concept to biomarker discovery in cancer research. In the third part of the dissertation, we explore the combined advantages of molecular interaction network and gene expression profiles to identify biomarkers in cancer research. The reliability of conventional gene biomarkers has been challenged because of the biological heterogeneity and noise within and across patients. In this dissertation, we present a module-based biomarker discovery approach that integrates interaction network topology and high-throughput gene expression data to identify markers not as individual genes but as modules. To select reliable biomarker sets across different studies, a hybrid method combining group feature selection with ensemble feature selection is proposed. First, a group feature selection method is used to extract the modules (subnetworks) with discriminative power between disease groups. Then, an ensemble feature selection method is used to select the optimal biomarker sets, in which a double-validation strategy is applied. The ensemble method allows combining features selected from multiple classifications with various data subsampling to increase the reliability and classification accuracy of the final selected biomarker set. The results from four breast cancer studies demonstrated the superiority of the module biomarkers identified by the proposed approach: they can achieve higher accuracies, and are more reliable in datasets with same clinical design. Based on the experimental results above, we believe that the proposed systems approaches provide meaningful solutions to discover the cellular regulatory processes and improve the understanding about disease mechanisms. These computational approaches are primarily developed for analysis of high-throughput genomic data. Nevertheless, the proposed methods can also be extended to analyze high-throughput data in proteomics and metablomics areas. / Ph. D.