Genomic microarray analysis is rapidly replacing conventional chromosome analysis by molecular karyotyping due to the significant increase in the power to detect causative CNVs. Here, we extensively validated the HumanHap550 and Human610-Quadv1_B Illumina platforms for potential diagnostic application by using patients with undiagnosed intellectual disability (ID). The first and foremost goal of our application study was to use these arrays for reliable genome wide detection of rare CNVs in patients of three different cohorts: 1) patients with unexplained intellectual disability 2) patients with unknown diffuse congenital hyperinsulinism (CHI) and 3) a family with a distinctive diagnosis of Holt-Oram syndrome (HOS). We showed that SNP-based arrays allow the detection of intragenic deletions and duplications. The identification of a disease-CNV affecting only a single gene allowed us to consider that particular gene as a candidate for intellectual disability. This was the case for three unrelated patients with moderate intellectual disability, global developmental delay, and severe speech and language disorders in which a de novo deletion encompassing solely the FOXP1 gene was detected. To prove further the causality of the FOXP1 deletion following-up investigations were based on a screening of the entire coding region of FOXP1 for nucleotide changes in a panel of 883 probands with intellectual disability. Eight non-synonymous coding changes, three synonymous and nine non-coding variants were identified. In addition to the de novo cases of ID, also patients suffering from an autosomal recessive form of ID were found in our cohort. We detected three partial heterozygous deletions of the COH1 gene at locus 8q22 which is mutated in Cohen syndrome. After sequencing the entire coding region and the exon/intron boundaries of COH1 we identified a stop mutation, a frameshift and two missense mutations in the remaining allele, respectively. Therefore, three compound heterozygous mutations were identified in the COH1 gene, thus providing a distinctive Cohen Syndrome diagnose to three unrelated patients of our ID cohort. We studied the genetic basis of a rare human autosomal disorder such as diffuse Congenital Hyperinsulinsm (CHI) in a cohort of 40 patients with inconspicuous mutation screening of ABCC8 and KCNJ11 genes. Chromosomal abnormalities detected by SNP oligonucleotide arrays accounted for 20% of the studied cases. The most interesting rearrangement was a 970kb deletion at the chromosomal band 1p31.1 which was found to encompass the PTGER3 and ZRANB2 genes and the last exon of the NEGR1 gene. We hypothesized that the haploinsufficiency of PTGER3 gene induces a 50% reduction of the stimulation by PGE2, thus diminishing the inhibition of glucose-stimulated insulin secretion (GSIS) and resulting in elevated insulin secretion. The screening for point mutations in the candidate gene PTGER3 did not reveal any pathogenic variant neither in the second allele of the patient in which a de novo deletion was detected nor in a cohort of 39 unrelated patients with unexplained CHI. Instead we identified a novel polymorphic variant which was also detected in 18 individuals of our control cohort. CNV analysis in a family with both atypical Holt-Oram syndrome and additional mammary glands was performed allowing the detection of a contiguous heterozygous duplication at the chromosomal band 12q24.21. The maximal duplication size could be estimated as aproximately 345,6kb including the whole coding region of the TBX5 and TBX3 genes. Gene dosage assessment at specific genetic loci demonstrated the cosegregation of the duplication and the Holt-Oram syndrome/supernumerary mammary glands phenotype in this pedigree, this being a strong indicator of its pathogenecity. Up to date, this is the first report of a heterozygous duplication encompassing both TBX5 and TBX3 genes, and consequently the first report of a combined phenotype of Holt-Oram syndrome and supernumerary mammary glands.

Genetic and genomic data often contain unobservable or missing information. Applications of probabilistic models such as mixture models and hidden Markov models (HMMs) have been widely used since the 1960s to make inference on unobserved information using some observed information demonstrating the versatility and importance of these models. Biological applications of mixture models include gene expression data, meta-analysis, disease mapping, epidemiology and pharmacology and applications of HMMs include gene finding, linkage analysis, phylogenetic analysis and identifying regions of identity-by-descent. An important statistical and informatics challenge posed by modern genetics is to understand the functional consequences of genetic variation and its relation to phenotypic variation. In the analysis of whole-exome sequencing data, predicting the impact of missense mutations on protein function is an important factor in identifying and determining the clinical importance of disease susceptibility mutations in the absence of independent data determining impact on disease. In addition to the interpretation, identifying co-inherited regions of related individuals with Mendelian disorders can further narrow the search for disease susceptibility mutations. In this thesis, we develop two probabilistic models in application of genetic and genomic data with missing information: 1) a mixture model to estimate a posterior probability of functionality of missense mutations and 2) a HMM to identify co-inherited regions in the exomes of related individuals. The first application combines functional predictions from available computational or {\it in silico} methods which often have a high degree of disagreement leading to conflicting results for the user to assess the pathogenic impact of missense mutations on protein function. The second application considers extensions of a first-order HMM to include conditional emission probabilities varying as a function of minor allele frequency and a second-order dependence structure between observed variant calls. We apply these models to whole-exome sequencing data and show how these models can be used to identify disease susceptibility mutations. As disease-gene identification projects increasingly use next-generation sequencing, the probabilistic models developed in this thesis help identify and associate relevant disease-causing mutations with human disorders. The purpose of this thesis is to demonstrate that probabilistic models can contribute to more accurate and dependable inference based on genetic and genomic data with missing information.

Brachyspira species are bacteria that colonize the intestines of some mammalian and avian species with different degrees of pathogenicity. Brachyspira intermedia is a mild pig and bird pathogen with an unknown genomic sequence. In this project, we completed the genome of Brachyspira intermedia PWS/AT and did a comparative genomic analysis between B. intermedia PWS/AT and the already completed genomes of B. hyodysenteriae WA1, B. murdochii 56-150T and B. pilosicoli 95/1000. A table containing 15 classes of unique and shared genes was developed and analyzed in order to gain a better understanding of species-specific traits and clues behind the different degree of pathogenicity. Our result shows that genes are overall poorly annotated and further studies are of great importance for understanding different and shared properties. The largest number of unique features was found in B. intermedia and B. murdochii. B. hyodysenteriae and B. pilosicoli has most likely developed independently towards different biological niches and B. pilosicoli has undergone a major reductive evolution. One plasmid and six prophages were found in B. intermedia, where two of the phages appear to be capable of horizontal gene transfer. Further genome sequencing of more strains will probably increase the understanding of species-specific traits even more.

<p>In the post-genomics era, the sheer volume of data is overwhelming without appropriate tools for data integration and analysis. Studying genomic sequences in the context of other related genomic sequences, i.e. comparative genomics, is a powerful technique enabling the identification of functionally interesting sequence regions based on the principal that similar sequences tend to be either homologous or provide similar functionality.</p> <p>Costs associated with full genome sequencing make it infeasible to sequence every genome of interest. Consequently, simple, smaller genomes are used as model organisms for more complex organisms, for instance, Mouse/Human. An annotated model organism provides a source of annotation for transcribed sequences and other gene regions of the more complex organism based on sequence homology. For example, the gene annotations from the model organism aid interpretation of expression studies in more complex organisms.</p> <p>To assist with comparative genomics research in the Arabidopsis/Brassica (Thale-cress/Canola) model-crop pair, a web-based, graphical genome browser (BioViz) was developed to display short Brassica genomic sequences in the context of the Arabidopsis model organism genome. This involved the development of graphical representations to integrate data from multiple sources and tools, and a novel user interface to provide the user with a more interactive web-based browsing experience. While BioViz was developed for the Arabidopsis/Brassica comparative genomics context, it could be applied to comparative browsing relative to other reference genomes.</p> <p>BioViz proved to be an valuable research support tool for Brassica / Arabidopsis comparative genomics. It provided convenient access to the underlying Arabidopsis annotation, allowed the user to view specific EST sequences in the context of the Arabidopsis genome and other related EST sequences. In addition, the limits to which the project pushed the SVG specification proved influential in the SVG community. The work done for BioViz inspired the definition of an opensource project to define standards for SVG based web applications and a standard framework for SVG based widget sets.</p>

Recent advances in DNA sequencing technology have enabled entire genomes to be sequenced quickly and accurately, resulting in an exponential increase in the number of organisms whose genome sequences have been elucidated. While the genome sequence of a given organism represents an important starting point in understanding its physiology, the functions of the protein products of many genes are still unknown; as such, computational methods for studying protein function are becoming increasingly important. In addition, this wealth of genomic information has created an unprecedented opportunity to compare the protein content of different organisms; among other applications, this can enable us to improve taxonomic classifications, to develop more accurate diagnostic tests for identifying particular bacteria, and to better understand protein content relationships in both closely-related and distantly-related organisms.<p> This thesis describes the design, evaluation, and use of a program called Proteome Subtraction and Intersection (PSI) that uses an idea called genome subtraction for discovering protein-phenotype relationships and for characterizing differences in protein content in groups of organisms. PSI takes as input a set of proteomes, as well as a partitioning of that set into a subset of "included" proteomes and a subset of "excluded" proteomes. Using reciprocal BLAST hits, PSI finds orthologous relationships among all the proteins in the proteomes from the original set, and then finds groups of orthologous proteins containing at least one orthologue from each of the proteomes in the "included" subset, and none from any of the proteomes in the "excluded" subset.<p> PSI is first applied to finding protein-phenotype relationships. By identifying proteins that are present in all sequenced isolates of the genus <i>Lactobacillus</i>, but not in the related bacterium <i>Pediococcus pentosaceus</i>, proteins are discovered that are likely to be responsible for the difference in cell shape between the lactobacilli and <i>P. pentosaceus</i>. In addition, proteins are identified that may be responsible for resistance to the antibiotic gatifloxacin in some lactic acid bacteria.<p> This thesis also explores the use of PSI for comparing protein content in groups of organisms. Based on the idea of genome subtraction, a novel metric is proposed for comparing the difference in protein content between two organisms. This metric is then used to create a phylogenetic tree for a large set of bacteria, which to the author's knowledge represents the largest phylogenetic tree created to date using protein content. In addition, PSI is used to find the proteomic cohesiveness of isolates of several bacterial species in order to support or refute their current taxonomic classifications.<p> Overall, PSI is a versatile tool with many interesting applications, and should become more and more valuable as additional genomic information becomes available.

Expansion of common bean (<i>Phaseolus vulgaris</i> L.) crops in the northern Great Planes has been hampered due to the lack of cultivars demonstrating sufficient vitality under low temperature conditions. <i>Phaseolus angustissimus</i> L., a wild bean species, has been previously shown to possess the ability to survive low temperatures in field trials. Freezing tolerance experiments under controlled conditions resulted in P. angustissimus demonstrating a greater capacity for freezing tolerance than <i>P. vulgaris</i>, as all P. vulgaris plants studied were dead at -2.5oC while most P. angustissimus plants treated to the same conditions survived. Exposure to chilling temperatures over five days resulted in stunted growth in both species, but the cultivated bean suffered more compared to the wild bean, as noted by a marked loss in tissue water content over the first three days of chilling. Interspecific macroarray hybridizations of a cDNA library from cold acclimated Medicago sativa L. using cDNAs derived from non-chilled and three-day chilled <i>P. vulgaris</i> and <i>P. angustissimus</i> plants showed that <i>P. vulgaris</i> showed more changes in gene expression after three days of chilling. Also, <i>P. vulgaris</i> showed a general trend towards down-regulation of the transcripts sampled on the third day of chilling compared to <i>P. angustissimus</i>. RT-PCR experiments were conducted using cDNAs from plant tissues exposed to various durations of chilling to confirm the results from the macroarray experiment. These time-course RT-PCR experiments revealed expression patterns across various chilling durations in genes identified from the macroarray. Data from these experiments suggest that <i>P. vulgaris</i> and <i>P. angustissimus </i> seedlings respond differently to low temperature exposure, and that some of the changes in <i>P. angustissimus</i> transcripts monitored here may be useful for researchers in better understanding how Phaseolus species can respond better to chilling temperatures.

The rapid development of large-scale genomic sequencing has dramatically changed the field of genetics, in part through the development of comparative genomics. Fungal comparative genomics is particularly powerful given the large number of genomes currently available, their compact architecture, and their relative ease of genetic manipulation. Fungal comparative genomics was employed in this work to address two related questions. First, it was used along with computational thermodynamic methods to predict non-coding RNA (ncRNA) in Saccharomyces cerevisiae. Sets of positive and negative control genes were evaluated to determine the effect of window sizes and step sizes on the sensitivity of ncRNA identification. The approach was then applied to predict ncRNA genes on chromosome 6 of S. cerevisiae and S. bayanus. Northern blot analysis, rapid amplification of cDNA ends (RACE), and publicly available cDNA library data were used to test the predictions. Strong experimental evidence was accumulated for four new ncRNA genes. Potential structural elements in the 5' and 3' untranslated regions (UTRs) of six annotated protein-coding genes were also identified. This work shows that thermodynamic approaches, coupled with comparative genomics, are powerful tools for predicting structural ncRNA. Second, comparative genomic approaches were employed to identify a non-reciprocal transfer event from Cryptococcus neoformans var. grubii to var. neoformans ~2 million years ago involving a 14 gene (~40 kb) region. The majority of clinical and environmental var. neoformans strains from around the world contain this sequence obtained from var. grubii. The introgression event likely occurred via an incomplete inter-varietal sexual cycle creating a hybrid intermediate where mobile elements common to both lineages mediated the exchange. The subsequent duplication in laboratory strains of a fragment of this same genomic region supports evolutionary theories that instabilities in subtelomeric regions promote adaptive evolution through gene amplification and subsequent adaptation. These data indicate that DNA exchange between closely related sympatric varieties or species may be a recurrent theme in the evolution of fungal species. It further suggests that while evolutionary divergence is the primary force driving speciation, rare introgression events also play a potentially important role. / Dissertation

<p>Allelic imbalance, in which two alleles at a given locus exhibit differences in gene expression, chromatin composition and/or protein binding, is a widespread phenomenon in the human and other complex genomes. Most examples concern individual loci located more or less randomly around the genome and thus imply local and gene-specific mechanisms. However, genomic or chromosomal basis for allelic imbalance is supported by multi-locus examples such as those exemplified by domains of imprinted genes, spanning ~1-2 Mb, or by X chromosome inactivation, involving much of an entire chromosome. Recent studies have shown that genes on the two female X chromosomes exhibit a breadth of expression patterns ranging from complete silencing of one allele to fully balanced biallelic expression. Although evidence for heritability of allele-specific chromatin and expression patterns exists at individual loci, it is unknown whether heritability is also reflected in the chromosome-wide patterns of X inactivation.</p><p>The aim of this thesis is to elucidate the extent to which the widespread variable patterns of allelic imbalance on the human X chromosome in females are under genetic control and how access of the transcription machinery to the human inactive X chromosome in females is determined at a genomic level. For the set of variable genes examined in this study, the absence or presence of expression appears to be stochastic with respect to the population rather than abiding by strict genetic rules. Furthermore, variable gene expression that I have detected even among multiple clonal cell lines derived from a single individual suggests fluctuation in transcriptional machinery engagement. I find that, although expression at most genes on the human inactive X chromosome is repressed as a result of X inactivation, a number of loci are accessible to the transcriptional machinery. It appears that RNA Polymerase II is present at alleles on the inactive X even at the promoters of several silenced genes, indicating a potential for expression. </p><p>This thesis embodies a transition in the field of human X chromosome inactivation from gene by gene approaches used in the past to utilizing high-throughput technologies and applying follow-up analytic techniques to draw upon the vast data publicly available from large consortia projects.</p> / Dissertation

MicroRNAs are a large family of 21-26 nucleotide non-coding RNAs with a role in the post-transcriptional regulation of gene expression. In recent years, microRNAs have been proposed to play a significant role in the expansion of organism complexity. MicroRNAs are expressed in a cell or tissue-specific manner during embryonic development, suggesting a role in cellular differentiation. For example, Let-7 is a metazoan microRNA that acts as developmental timer between larval stages in C. elegans. We conducted a comparative study that determined the distribution of microRNA families among metazoans, including the identification of new family members for several species. MicroRNA families appear to have evolved in bursts of evolution that correlate with the advent of major metazoan groups such as vertebrates, eutherians, primates and hominids. Most microRNA families identified in these organisms appeared with or after the advent of vertebrates. Only a few of them appear to be shared between vertebrates and invertebrates. The distribution of these microRNA families supports the idea that at least one whole genome duplication event (WGS) predates the advent of vertebrates. Gene ontology analyses of the genes these microRNA families regulate show enrichments for functions related to cell differentiation and morphogenesis. MicroRNA genes appear to be under great selective constraints. Identification of conserved regions by comparative genomics allows for the computational identification of microRNAs. We have identified and characterized ultraconserved regions between the genomes of the honey bee (Apis mellifera) and the parasitic wasp (Nasonia vitripennis), and developed a strategy for the identification of microRNAs based on regions of ultraconservation. Ultraconserved regions preferentially localize within introns and intergenic regions, and are enriched in functions related to neural development. Introns harboring ultraconserved elements appear to be under negative selection and under a level of constraint that is higher than in their exonic counterparts. This level of constraint suggests functional roles yet to be discovered and suggests that introns are major players in the regulation of biological processes. Our computational strategy was able to identify new microRNA genes shared between honey bee and wasp. We recovered 41 of 45 previously validated microRNAs for these organisms, and we identified several new ones. A significant fraction of these microRNA candidates are located in introns and intergenic regions and are organized in genomic clusters. Expression of 13 of these new candidates was verified by 454 sequencing.

Although the completion of the Human Genome Project helps develop efficient treatment/prevention programs, it will raise new and non-trivial public health issues. Many of these issues fall under the professional purview of health educators. Yet, no studies have evaluated if health educators (HEs) are ready to adopt genomic competencies into health promotion. This dissertation addresses this issue by examining three research questions in three separate studies: 1) Why must HEs develop genomic competencies? 2) What are HEs’ knowledge of, and attitudes toward genomic competencies? And 3) what is HEs’ likelihood of adopting genomic competencies into health promotion? The first theoretical study proposed five arguments supporting the need for HEs to develop their genomic competencies and integrate public health genomics into health promotion. These arguments touched on various dimensions of HEs’ professional goals and ranged from professional responsibilities and competencies, to the availability of funding for genomic-related research or interventions and opportunities for future employment. For the second study, a web-based survey was developed and distributed to all members of four major health education organizations. A total of 1,925 HEs’ completed the survey and 1,607 responses were utilized in the final analysis. This study indicated that participants had deficient knowledge and unfavorable attitudes toward the CDCproposed genomic competencies. In the third study, a theoretical model was developed to predict HEs’ likelihood to incorporate genomic competencies into their practice. Using techniques from Structural Equation Modeling (SEM), the model was tested with the same data of the second study. Findings supported the proposed theoretical model. While genomic knowledge, attitudes, and self-efficacy were significantly associated with HEs’ likelihood to incorporate genomic competencies into their practice, attitudes was the strongest predictor of likelihood. In summary, these studies indicated that participating HEs had deficient genomic knowledge, unfavorable attitudes toward a set of CDC-proposed genomic competencies, and low likelihood to adopt genomic competencies into health promotion. Relevant training should be developed and advocated. As the SEM analysis results indicated the survey findings supported the proposed theoretical model, which can be utilized to steer future training for HEs.