Submitted to the Faculty of the School of Informatics in parial fulfillment of the requirements for the degree of Master of Schience in Bioinformatics Indiana University August 2006 / Decoding the human genome resulted in generating large amount of data that need to be analyzed and given a biological meaning. The field of Life Schiences is highly information driven. The genomic data are mainly the gene expression data that are obtained from measurement of mRNA levels in an organism. Efficiently processing large amount of gene expression data has been possible with the help of high throughput technology. Research studies working on microarray data has led to the possibility of finding disease biomarkers. Carrying out biomarker discovery experiments has been greatly facilitated with the emergence of various analytical and visualization tools as well as annotation databases. These tools and databases are often termed as 'bioinformatics services'. The main purpose of this research was to develop SIBIOS (Bystem for Integration of Bioinformatics Services) as a platform to carry out microarray experiments for the purpose of biomarker discovery. Such experiments require the understanding of the current procedures adopted by researchers to extract biologically significant genes. In the course of this study, sample protocols were built for the purpose of biomarker discovery. A case study on the BCR-ABL subtype of ALL was selected to validate the results. Different approaches for biomarker discovery were explored and both statistical and mining techniques were considered. Biological annotation of the results was also carried out. The final task was to incorporate the new proposed sample protocols into SIBIOS by providing the workflow capabilities and therefore enhancing the system's characteristics to be able to support biomarker discovery workflows.

京都大学 / 0048 / 新制・論文博士 / 博士(医学) / 乙第13157号 / 論医博第2144号 / 新制||医||1028(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 影山 龍一郎, 教授 別所 和久, 教授 辻川 明孝 / 学位規則第4条第2項該当 / Doctor of Medical Science / Kyoto University / DFAM

No description available.

No description available.

No description available.

Advancements in DNA microarray data sequencing have created the need for sophisticated machine learning algorithms and feature selection methods. Probabilistic graphical models, in particular, have been used to identify whether microarrays or genes cluster together in groups of individuals having a similar diagnosis. These clusters of genes are informative, but can be misleading when every gene is used in the calculation. First feature reduction techniques are explored, however the size and nature of the data prevents traditional techniques from working efficiently. Our method is to use the partial correlations between the features to create a precision matrix and predict which associations between genes are most important to predicting Leukemia diagnosis. This technique reduces the number of genes to a fraction of the original. In this approach, partial correlations are then extended into a spectral clustering approach. In particular, a variety of different Laplacian matrices are generated from the network of connections between features, and each implies a graphical network model of gene interconnectivity. Various edge and vertex weighted Laplacians are considered and compared against each other in a probabilistic graphical modeling approach. The resulting multivariate Gaussian distributed clusters are subsequently analyzed to determine which genes are activated in a patient with Leukemia. Finally, the results of this are compared against other feature engineering approaches to assess its accuracy on the Leukemia data set. The initial results show the partial correlation approach of feature selection predicts the diagnosis of a Leukemia patient with almost the same accuracy as using a machine learning algorithm on the full set of genes. More calculations of the precision matrix are needed to ensure the set of most important genes is correct. Additionally more machine learning algorithms will be implemented using the full and reduced data sets to further validate the current prediction accuracy of the partial correlation method.

Microarrays measure the abundance of DNA transcripts for thousands of gene sequences, simultaneously facilitating genomic comparisons across tissue types or disease status. These experiments are used to understand fundamental aspects of growth and development and to explore the underlying genetic causes of many diseases. The data from most microarray studies are found in open-access online databases. Bayesian models are ideal for the analysis of microarray data because of their ability to integrate prior information; however, most current Bayesian analyses use empirical or flat priors. We present a Perl script to build an informative prior by mining online databases for similar microarray experiments. Four prior distributions are investigated: a power prior including information from multiple previous experiments, an informative prior using information from one previous experiment, an empirically estimated prior, and a flat prior. The method is illustrated with a two-sample experiment to determine the preferential regulation of genes by tamoxifen in breast cancer cells.

Understanding the role of deregulated biological pathways in cancer samples has the potential to improve cancer treatment, making it more effective by selecting treatments that reverse the biological cause of the cancer. One of the challenges with pathway analysis is identifying a deregulated pathway in a given sample. This project develops the Universal Probability of Expression Code (UPC), a profile of a single deregulated biological path- way, and projects it into a cancer cell to determine if it is present. One of the benefits of this method is that rather than use information from a single over-expressed gene, it pro- vides a profile of multiple genes, which has been shown by Sjoblom et al. (2006) and Wood et al. (2007) to be more effective. The UPC uses a novel normalization and summarization approach to characterize a deregulated pathway using only data from the array (Mixture model-based analysis of expression arrays, MMAX), making it applicable to all microarray platforms, unlike other methods. When compared to both Affymetrix's PMA calls (Hubbell, Liu, and Mei 2002) and Barcoding (Zilliox and Irizarry 2007), it performs comparably.

DNA microarrays are a potentially disruptive technology in the medical field, but their use in such settings is limited by poor reliability. Microarrays work on the principle of hybridization and can only be as reliable as this process is robust, yet little is known at the molecular level about how the surface affects the hybridization process. This work uses advanced molecular simulation techniques and an experimentally-parameterized coarse-grain model to determine the mechanism by which hybridization occurs on surfaces and to identify key factors that influence the accuracy of DNA microarrays. Comparing behavior in the bulk and on the surface showed, contrary to previous assumptions, that hybridization on surfaces is more energetically favorable than in the bulk. The results also show that hybridization proceeds through a mechanism where the untethered (target) strand often flips orientation. For evenly-lengthed strands, the surface stabilizes hybridization (compared to the bulk system) by reducing the barriers involved in the flipping event. Additional factors were also investigated, including the effects of stretching or compressing the probe strand as a model system to test the hypothesis that improving surface hybridization will improve microarray performance. The results in this regard indicate that selectivity can be increased by reducing overall sensitivity by a small degree. Another factor that was investigated was the effect of unevenly-lengthed strands. It was found that, when unevenly-lengthed strands were hybridized on a surface, the surface may destabilize hybridization compared to the bulk, but the degree of destabilization is dependent on the location of the matching sequence. Taken as a whole, the results offer an unprecedented view into the hybridization process on surfaces and provide some insights as to the poor reproducibility exhibited by microarrays. Namely, the prediction methods that are currently used to design microarrays based on duplex stability in the bulk do a poor job of estimating the stability of those duplexes in a microarray environment.

Gene regulation is controlled by DNA-bound complexes of transcription factors (TFs) and indirectly recruited transcriptional cofactors (COFs). Understanding how and where these TF-COF complexes bind in the genome is fundamental to our understanding of the role of cis-regulatory elements (CREs) in gene regulation and our mechanistic interpretation of non-coding variants (NCVs) known to impact gene expression levels. In this thesis, I present three related array-based techniques for the high-throughput profiling of DNA-bound TFs and TF-COF complexes directly from cell nuclear extracts. First, I describe the nuclear extract protein-binding microarray (nextPBM) approach to profile TF-DNA binding using nuclear extracts to account for cell-specific post-translational modifications and cofactors. By analyzing cooperative binding of PU.1/SPI1 and IRF8 in monocytes, I demonstrate how nextPBM can be used to delineate DNA-sequence determinants of cell-specific cooperative TF complexes. Second, I present the CASCADE (Comprehensive ASsessment of Complex Assembly at DNA Elements) approach to simultaneously discover DNA-bound TF-COF complexes and quantify the impact of NCVs on their binding. To demonstrate applicability of CASCADE to screen NCVs, I profile differential TF-COF binding to ~1,700 single-nucleotide polymorphisms in human macrophages and discover a prevalence of perturbed ETS-related TF-COF complexes at these quantitative trait loci. Third, I present the human TF array (hTF array) as a general platform for surveying COF recruitment to a panel of 346 non-redundant consensus TF binding sites (TFBSs). Using the hTF array, one can examine the activity of a diverse panel of TFs by profiling TF-COF complexes in a cell state-specific manner. In addition to the hTF microarray design, I have developed analysis and visualization software that allows users to explore COF recruitment profiling results interactively. Collectively, nextPBM, CASCADE, and the hTF array represent a suite of new approaches to investigate TF-COF complex binding and their application will refine our understanding of CREs by linking NCVs with the biophysical complexes that mediate gene regulatory functions.