Genetic data analysis for classification of bipolar disordersGenetic data analysis for classification of bipolar disorders Διπλωματική Εργασία Diploma Work 2015-09-252015enIn the resent years DNA microarray analysis has become a widely used tool for gene expression profiling and data analysis. This technology can be useful in the classification of complex diseases such as bipolar disorder, providing useful information for its genetic background. Bipolar disorder is a common, heritable mental illness characterized by recurrent episodes of mania and depression that manifests from multiple genetic and environmental factors. There are four basic types of bipolar disorder; bipolar I disorder, dipolar II disorder, Bipolar Disorder Not Otherwise Specified (BP-NOS) and Cyclothymic. The ability to classify dipolar disorders may have a major impact on our understanding of disease pathophysiology and may provide important opportunities to investigate the interaction between genetic and environmental factors involved in pathogenesis. Also this ability may be essential to guide appropriate therapy and determine prognosis for successful treatment. The aim of this diploma thesis is to extract a significant genomic signature for which biological knowledge already exists or discover novel genomic information, which might stand as the motivation for further analysis. Under this genomic signature we classify the bipolar disorders using gene expressions from two different populations. Microarray analysis normally leads to datasets which contain a small number of samples which have a large number of gene expression levels as features. In order to extract useful informative sets of genes that can reduce dimensionality and maximize the performance of classifiers, feature selection algorithms were used. Another aim of this study is to achieve stable performance assessment of feature selection and classification methods. In that manner, the genetic evaluation framework named “Stable Bootstrap Validation” (SBV), introduced be Nick Chlis, is presented. The SBV utilizes bootstrap resampling of the original dataset and an explicit criterion that determines the stability of the observed classification accuracy and the biological interpretation of genes, also called genomic signature. Moreover, methodologies for evaluating the discrimination, consistency and generalization ability of the observed results are also introduced. In this diploma thesis a unified “32 common gene signature” was extracted, which is closely associated with several aspect of bipolar disorders. http://creativecommons.org/licenses/by/4.0/Πολυτεχνείο Κρήτης::Σχολή Ηλεκτρονικών Μηχανικών και Μηχανικών ΥπολογιστώνValsamo_Leska_Dip_2015.pdfChania [Greece]Library of TUC2015-09-25application/pdf3.5 MBfree Leska Valsamo Λεσκα Βαλσαμω Zervakis Michalis Ζερβακης Μιχαλης Balas Costas Μπαλας Κωστας Petrakis Evripidis Πετρακης Ευριπιδης Technical University of Crete Πολυτεχνείο Κρήτης Bipolar depression Bipolar disorder Depression, Bipolar Depression, Manic Manic depression Manic-depressive psychoses Manic-depressive psychosis Melancholia manic depressive illness bipolar depression bipolar disorder depression bipolar depression manic manic depression manic depressive psychoses manic depressive psychosis melancholia Bio-informatics Biological informatics bioinformatics bio informatics biological informatics