Breast cancers with overexpression or amplification of the HER2 tyrosine kinase receptor are more aggressive, resistant to chemotherapy, and associated with a worse prognosis. Currently, these breast cancers are treated with the monoclonal antibody trastuzumab (Herceptin®). Unfortunately, not all patients respond to trastuzumab drug therapy; some patients show de novo resistance, while others acquire resistance during treatment. This thesis describes our RNAi studies to identify novel regulators of the HER2 signaling pathway in breast cancer. Three kinome-wide siRNA screens were performed on five HER2 amplified and seven HER2 non-amplified breast cancer cell lines, two normal breast cell lines, as well as two HER2-positive breast cancer cell lines with acquired trastuzumab resistance and their isogenic trastuzumab-sensitive controls. To understand the main kinase drivers of HER2 signaling, we performed a comprehensive screen that selected against growth inhibitors of the non-HER2 amplified breast cancer cell lines. This screen identified the loss of the HER2/HER3 heterodimer as the most prominent selective inhibitor of HER2-amplified breast cancers. In a trastuzumab sensitization screen on five trastuzumab-treated breast cancer cell lines, we identified several siRNA against the PI3K pathway as well as various other signaling pathways that inhibited proliferation. Finally, in a screen for acquired trastuzumab resistance, PKCη and its downstream targets were identified. Loss of PKCη resulted in a decrease in G1/S transition and upregulation of the cyclin dependent kinase inhibitor p27. Initial data suggest that PKCη promotes p27 ubiquitination and degradation. Taken together, these studies provide novel insight into the complex signaling of HER2-positive breast cancers and the mechanisms of resistance to trastuzumab therapy. This work describes how various kinases can modulate cell proliferation, and points to possible novel drug targets for the treatment of HER2-positive breast cancers.

Glycosylation, the addition of one or more carbohydrates molecules to a protein, is crucial for many cellular processes. Aberrant glycosylation is a key marker for various diseases such as cancer and rheumatoid arthritis. It has also recently been discovered that glycosylation is important in the ability of the Human Immunodeficiency Virus (HIV) to evade recognition by the immune system. Given the importance of glycosylation in disease, major efforts are underway in life science research to investigate the glycome, the entire glycosylation profile of an organelle, cell or tissue type. To date, little bioinformatics research has been performed in glycomics due to the complexity of glycan structures and the low throughput of carbohydrate analysis. Recent advances in mass spectrometry (MS) have greatly facilitated the analysis of the glycome. Increasingly, this technology is preferred over traditional methods of carbohydrate analysis which are often laborious and unsuitable for low abundance glycoproteins. When subject to mass spectrometry with collision-induced dissociation, glycopeptides produce characteristic MS/MS spectra that can be detected by visual inspection. However, given the high volume of data output from proteome studies today, manually searching for glycopeptides is an impractical task. In this thesis, we present a tool to automate the identification of glycopeptide spectra from MS/MS data. Further, we discuss some methodologies to automate the elucidation of the structure of the carbohydrate moiety of glycopeptides by adapting traditional MS/MS ion searching techniques employed in peptide sequence determination. MS/MS ion searching, a common technique in proteomics, aims to interpret MS/MS spectra by correlating structures from a database to the patterns represented in the spectrum. The tool was tested on high throughput proteomics data and was shown to identify 97% of all glycopeptides present in the test data. Further, the tool assigned correct carbohydrate structures to many of these glycopeptide MS/MS spectra. Applications of the tool in a proteomics environment for the analysis of glycopeptide expression in cancer tissue are also be presented.

The term of Cognitive Radio (CR) used to indicate that spectrum radio could be accessed dynamically and opportunistically by unlicensed users. In CR Networks, Interference between nodes, hidden terminal problem, and spectrum sensing errors are big issues to be widely discussed in the research field nowadays. To improve the performance of such kind of networks, this thesis proposes Cognitive Medium Access Control (MAC) protocols for Mobile Ad-Hoc Networks (MANETs). From the concept of CR, this thesis has been able to develop a cognitive MAC framework in which a cognitive process consisting of cognitive elements is considered, which can make efficient decisions to optimise the CR network. In this context, three different scenarios to maximize the secondary user's throughput have been proposed. We found that the throughput improvement depends on the transition probabilities. However, considering the past information state of the spectrum can dramatically increases the secondary user's throughput by up to 40%. Moreover, by increasing the number of channels, the throughput of the network can be improved about 25%. Furthermore, to study the impact of Physical (PHY) Layer errors on cognitive MAC layer in MANETs, in this thesis, a Sensing Error-Aware MAC protocols for MANETs has been proposed. The developed model has been able to improve the MAC layer performance under the challenge of sensing errors. In this context, the proposed model examined two sensing error probabilities: the false alarm probability and the missed detection probability. The simulation results have shown that both probabilities could be adapted to maintain the false alarm probability at certain values to achieve good results. Finally, in this thesis, a cooperative sensing scheme with interference mitigation for Cognitive Wireless Mesh Networks (CogMesh) has been proposed. Moreover, a prioritybased traffic scenario to analyze the problem of packet delay and a novel technique for dynamic channel allocation in CogMesh is presented. Considering each channel in the system as a sub-server, the average delay of the users' packets is reduced and the cooperative sensing scenario dramatically increases the network throughput 50% more as the number of arrival rate is increased.

Stream programming is a paradigm where a program is structured by a set of computational nodes connected by streams. Focusing on data moving between computational nodes via streams, this programming model fits well for applications that process long sequences of data. We call such applications reactive stream programs (RSPs) to distinguish them from stream programs with rather small and finite input data. In stream programming, concurrency is expressed implicitly via communication streams. This helps to reduce the complexity of parallel programming. For this reason, stream programming has gained popularity as a programming model for parallel platforms. However, it is also challenging to analyse and improve the performance without an understanding of the program's internal behaviour. This thesis targets an effi cient execution model for deploying RSPs on parallel platforms. This execution model includes a monitoring framework to understand the internal behaviour of RSPs, scheduling strategies for RSPs on uniform shared-memory platforms; and mapping techniques for deploying RSPs on heterogeneous distributed platforms. The foundation of the execution model is based on a study of the performance of RSPs in terms of throughput and latency. This study includes quantitative formulae for throughput and latency; and the identification of factors that influence these performance metrics. Based on the study of RSP performance, this thesis exploits characteristics of RSPs to derive effective scheduling strategies on uniform shared-memory platforms. Aiming to optimise both throughput and latency, these scheduling strategies are implemented in two heuristic-based schedulers. Both of them are designed to be centralised to provide load balancing for RSPs with dynamic behaviour as well as dynamic structures. The first one uses the notion of positive and negative data demands on each stream to determine the scheduling priorities. This scheduler is independent from the runtime system. The second one requires the runtime system to provide the position information for each computational node in the RSP; and uses that to decide the scheduling priorities. Our experiments show that both schedulers provides similar performance while being significantly better than a reference implementation without dynamic load balancing. Also based on the study of RSP performance, we present in this thesis two new heuristic partitioning algorithms which are used to map RSPs onto heterogeneous distributed platforms. These are Kernighan-Lin Adaptation (KLA) and Congestion Avoidance (CA), where the main objective is to optimise the throughput. This is a multi-parameter optimisation problem where existing graph partitioning algorithms are not applicable. Compared to the generic meta-heuristic Simulated Annealing algorithm, both proposed algorithms achieve equally good or better results. KLA is faster for small benchmarks while slower for large ones. In contrast, CA is always orders of magnitudes faster even for very large benchmarks.

Master of Science / Department of Computing and Information Sciences / Mitchell L. Neilsen / This research is a part of BREAD PHENO, a PhenoApps BREAD project at K-State which combines contemporary advances in image processing and machine vision to deliver transformative mobile applications through established breeder networks. In this platform, novel image analysis segmentation algorithms are being developed to model and extract plant phenotypes. As a part of this research, the traditional Watershed segmentation algorithm has been extended and the primary goal is to accurately count and characterize the seeds in an image. The new approach can be used to characterize a wide variety of crops. Further, this algorithm is migrated into Android making use of the Android APIs and the first ever user-friendly Android application implementing the extended Watershed algorithm has been developed for Mobile field-based high-throughput phenotyping (HTP).

No description available.

No description available.

In vitro protein selection has had major impacts in the field of protein engineering. Traditional screens assay individual proteins for specific function. Selection, however, analyzes a pool of mutants and yields the best variants. Phage display, a successful selection technique, also provides a reliable link between variant phenotype and genotype. It can also be coupled with high throughput sequencing to map protein mutations; potentially highlighting vital mutations in variants. We propose to apply this technique to cancer therapy. RAF, a serine/threonine kinase, is critical for cell regulation in mammals. RAF can be activated by oncogenic RAS, found in over 30% of cancers, to drive cancer proliferation. Rigosertib, a benzyl styryl sulfone in phase III clinical trials for myelodysplastic syndrome (MDS), is an inhibitor of the RAS binding domain (RBD) in RAF. Phage display can be used to select RAF mutants for RAS binding affinity, in the presence of Rigosertib. High-throughput sequencing of these variants can provide a means of anticipating, and mapping resistance to this anti-cancer drug.

In this thesis, a highly comparative framework for time-series analysis is developed. The approach draws on large, interdisciplinary collections of over 9000 time-series analysis methods, or operations, and over 30 000 time series, which we have assembled. Statistical learning methods were used to analyze structure in the set of operations applied to the time series, allowing us to relate different types of scientific methods to one another, and to investigate redundancy across them. An analogous process applied to the data allowed different types of time series to be linked based on their properties, and in particular to connect time series generated by theoretical models with those measured from relevant real-world systems. In the remainder of the thesis, methods for addressing specific problems in time-series analysis are presented that use our diverse collection of operations to represent time series in terms of their measured properties. The broad utility of this highly comparative approach is demonstrated using various case studies, including the discrimination of pathological heart beat series, classification of Parkinsonian phonemes, estimation of the scaling exponent of self-affine time series, prediction of cord pH from fetal heart rates recorded during labor, and the assignment of emotional content to speech recordings. Our methods are also applied to labeled datasets of short time-series patterns studied in temporal data mining, where our feature-based approach exhibits benefits over conventional time-domain classifiers. Lastly, a feature-based dimensionality reduction framework is developed that links dependencies measured between operations to the number of free parameters in a time-series model that could be used to generate a time-series dataset.

The homologous recombination (HR) pathway of DNA repair is essential for the faithful repair of double-stranded DNA breaks (DSBs) in all organisms and as such helps maintain genomic stability. Furthermore, HR is instrumental in the cellular response to exogenous DNA damaging agents such as those used in the clinic for chemo- and radiotherapy. HR in humans is a complex, incompletely understood process involving numerous stages and diverse biochemical activities. Advancing our knowledge of the HR pathway in humans aids the understanding of how chemo- and radiotherapies act and may be used to develop novel therapeutic strategies. Recent studies have identified inhibition of HR as one of the mechanisms via which a number of recently developed chemotherapeutics have their effect. Accordingly, the clinical potential of HR inhibitors is under investigation. My work has centred around the identification of both novel HR proteins and novel, small molecule HR inhibitors. To further these aims, I have successfully employed high-throughput RNAi and small molecule screening strategies. RNAi screens are commonly used to identify genes involved in a given cellular process via genetic loss of function, whilst small molecule, cell based screens are a powerful tool in the drug discovery process.