In this dissertation, we describe a new approach for gene finding that can utilize proteomics information in addition to DNA and RNA to identify new genes in prokaryote genomes. Proteomics processing pipelines require identification of small pieces of proteins called peptides. Peptide identification is a very error-prone process and we have developed a new algorithm for validating peptide identifications using a distance-based outlier detection method. We demonstrate that our method identifies more peptides than other popular methods using standard mixtures of known proteins. In addition, our algorithm provides a much more accurate estimate of the false discovery rate than other methods. Once peptides have been identified and validated, we use a second algorithm, proteogenomic mapping (PGM) to map these peptides to the genome to find the genetic signals that allow us to identify potential novel protein coding genes called expressed Protein Sequence Tags (ePSTs). We then collect and combine evidence for ePSTs we generated, and evaluate the likelihood that each ePST represents a true new protein coding gene using supervised machine learning techniques. We use machine learning approaches to evaluate the likelihood that the ePSTs represent new genes. Finally, we have developed new approaches to Bayesian learning that allow us to model the knowledge domain from sparse biological datasets. We have developed two new bootstrap approaches that utilize resampling to build networks with the most robust features that reoccur in many networks. These bootstrap methods yield improved prediction accuracy. We have also developed an unsupervised Bayesian network structure learning method that can be used when training data is not available or when labels may not be reliable.

High throughput technologies such as microarrays and modern genome sequencers produce enormous amounts of data that require novel data processing. This thesis proposes a method called Interdependent Correlation Cluster (ICC) to analyze the relations between genes represented by microarray data that are conditioned on a specific target gene. Based on Correlation Clustering, the proposed method analyzes a large set of correlation values related to the gene expression profiles extracted from given microarray datasets. The proposed method works on any size microarray datasets and could be applied to any target gene. In this study the selected target gene, NSF1 /USV1 / YPL230W, encodes a poorly characterized C2H2 zinc finger transcription factor (TF) involved in stress responses in yeast. The method is successful in the identification of novel NSF1 functional roles during fermentation stress conditions in the M2 industrial yeast strain. The new identified functions include regulation of energy and sulfur metabolism, protein synthesis, ribosomal assembly and protein trafficking as well as other processes. NSF1 involvement in sulfur metabolism was experimentally confirmed using biological laboratory techniques. Importantly, implication of NSF1 in sulfur metabolism regulation has highly relevant implications to wine and beer production industries concerned with production of compounds having sulfur-like off odour (SLO) and toxic properties. The correlation clustering also provides a means of understanding complex interactions existing between genes. / The pdf file contains numerous hyperlinks and bookmarks to facilitate navigation. This thesis will be of interest to those working with topics such as data mining of microarray data, novel gene function discovery and prediction, and genome-wide responses to fermentation stresses. / Ministry of Training, Colleges and Universities of Ontario (Ontario Graduate Scholarship and Ontario Graduate Scholarships in Science and Technology); The Natural Sciences and Engineering Research Council of Canada (NSERC)

The congenital myasthenic syndromes (CMS) are a group of rare genetic diseases affecting the neuromuscular junction (NMJ). These syndromes affect signal transmission and result in fatigable muscle weakness. In this study we performed exome analysis of 2 CMS patient cohorts and identified SYT11, a synaptotagmin inhibitor of clathrin mediated endocytosis (CME), and MGAT5B, a glycosylation protein, as potential novel CMS genes using bioinformatic analysis on the RD-Connect Genome Phenome Analysis Platform (GPAP). To validate them, we utilized morpholino knockdown models of zebrafish orthologues syt11a, syt11b, and mgat5b and conducted functional assays measuring chorion activity and escape response. Our results show that co-knockdown of syt11a/b or syt11b alone (and not mgat5b) results in a substantial neuromuscular deficit, with ablation of chorion activity and severely reduced escape response. Immunofluorescent studies showed both motor neuron growth and NMJ formation was inhibited by syt11a/b knockdown. In conclusion, syt11b causes a severe neuromuscular phenotype in zebrafish which supports SYT11 as a novel CMS-causing gene.

Placenta is a transient association of the fetal and maternal tissues, that develops during pregnancy, in most viviparous animals. The evolution of placenta ensured the development of the fetus inside the womb of the mother, providing a protected environment for the development of the fetus, and preventing the loss of progeny due to unfavorable environmental conditions. Because it is strategically poised at the maternal and fetal interface, the placenta is ideally suited to carry out alimentary, respiratory and excretory functions for the developing fetus. In addition, it serves as an immunological barrier preventing the rejection of the fetal semi-allograft, by the maternal immune system. Furthermore, the placenta elaborates a variety of protein, polypeptide and steroid hormones. These include growth factors, growth factor receptors, neuropeptides, opioids, progesterone and estrogen, whose secretion is dependent on the gestational age of the placenta and its differentiation status. The human placenta, adapts itself remarkably to cater to the changing requirements of the developing fetus. For instance, during the first trimester of pregnancy, the placenta is an actively dividing, a highly invasive and a rapidly differentiating organ; while near term, it represents a fully differentiated and a non-invasive unit. Furthermore, the placenta of the first trimester and that at term differ in their hormone profiles, extents of apoptosis, expression of several transcription factors, etc. This dramatic change in the phenotype of the human placenta can be considered to be the outcome of an intrinsically programmed pattern of differentiation, which may be referred to as the functional differentiation of the placenta. It may be hypothesized therefore, that this functional differentiation could be brought about by the differential expression of genes in the first trimester and the term placenta. The objectives of the present study were: 1. To gain an insight into this process of " functional differentiation” by investigating the differential expression of genes in the two developmentally distinct stages during gestation, viz. during the first trimester and at term. 2. To understand the functional relevance of the differentially expressed genes. A general introduction of the human placenta, describing the importance of differential expression in modulating placental function, is discussed in chapter 1. The functions of the human placenta along with a brief description of its development and differentiation are also briefly described. A Differential Display RT-PCR-based (DD RT-PCR) approach, using total RNA from the first-trimester and term placental villi, was employed to display the differentially expressed genes in the first trimester and the term placenta. The display so generated was used to identify a few differentially expressed cDNAs. This study was aimed at understanding the functional significance of the transcripts which were identified from the display, rather than just concentrate on documenting the differences in the gene expression patterns in the first trimester and the term placental tissue. A detailed description of the methodology adopted for performing DD-PCR using placental tissue, discussing the advantages and disadvantages of using differential display PCR, is described in chapter2. The use of DD-PCR for studying differential gene expression in the human placenta was validated by the finding that one of the cDNAs that was differentially expressed in the first trimester placental tissue, is a fragment of β-hCG cDNA. It is well documented that the differential expression of the β-subunit of hCG (human chorionic gondatrophin) during the first eight weeks of gestation is the rate limiting step in the synthesis and secretion of the functional hormone, which comprises the α and the β-subunits. Furthermore, the use of the model system viz., the first trimester and term placental tissue, was also validated for carrying out DD-PCR by ensuring that all placental samples used for DD analysis were free of endometrial contamination. A detailed description of optimization and validation of DD-PCR in human placental tissues is given in chapter 2. Cloning and sequencing of yet another cDNA from the first trimester differential display revealed that it is T-Plastin. T-Plastin is a member of a family of proteins that are involved in actin-bundling. Northern blot analysis and immunohistochemical studies using an antibody generated to a peptide corresponding to human T-Plastin, confirmed its differential expression and localization in the first trimester placenta. Considering the fact that several carcinomas show enhanced expression of T-Plastin, we tested the hypothesis that its differential expression is correlated with the proliferative potential of the first-trimester placenta It was observed that the first-trimester tissue expressed high levels of beta-actin as compared to the term placental tissue. This is in agreement with the up-regulation of beta-actin following mitogenic stimulation/proliferation and during neoplastic transformation or transformation-associated invasive behaviour of cells, two characteristic features shared by the early placenta with cancerous tissues. Based on our studies and available information in the literature, it is proposed that T-Plastin expression in the first trimester placenta is a growth-associated phenomenon which is partially responsible for the tumor-like phenotype of the first trimester tissue. Studies carried out with the partial T-Plastin cDNA clone that was isolated from the first trimester differential display, are presented in chapter 3. Sequencing of yet another cDNA clone identified from the term placental differential display, T-18 revealed that it had no homology to any known sequence in the nucleotide or est databases. The sequence corresponding to this clone was submitted to the GenBank and was assigned an accession number- AF089811. The differential expression of T-18 was confirmed by Northern blot analysis and RT-PCR analysis. Attempts were made to isolate the full-length cDNA corresponding to T-18 from a commercially available library from Clontech. However, repeated trials to identify the clone corresponding to T-18 did not yield any positive results. However, a genome database search revealed that T-18 was a portion of a large contig contained in chromosome 15. Analysis of the annotated gene sequences in and around the region in which T-18 is located in chromosome 15, revealed that there are very few ests reported in this contig and quite a few repeat sequences reported. Interestingly, it was observed that 6 kb downstream of the region in which T-18 is located, there was an est that had homology to a Bcl-2 precursor protein (an evolutionarily conserved, anti-apoptotic protein, capable of conferring protection against death-inducing signals) and the death adaptor protein, CRADD {Caspase and RIP adapter with death domain). Further updating of the ests in the database might probably be of help in the identification of the full-length cDNA corresponding to T-18 and confirm as to whether T-18 is a part of the gene/gene cluster that comprises the afore-mentioned est. An account of the identification and cloning of T-18 from the term placenta and the attempts to isolate the full-length cDNA clone corresponding to T-18 from a term placental cDNA library, is described in chapter 4. In the absence of any information on the identity of T-18, a study to understand the functional significance of T-18 expression was carried out. Since it was not possible to carry out studies pertaining to the temporal expression of T-18 throughout gestation on the human placenta for ethical reasons, alternate animal/organ models were employed to study T-18 expression. Rat placenta and rat Corpus Luteum (CL) were chosen as alternate models for studying T-18 expression as these two organs/tissues underwent dynamic changes in their function throughout pregnancy. For instance, it is well known that CL is the primary source of progesterone for maintaining pregnancy in the rat and that the progesterone secreting capacity of the luteal cells peak on day 16 of gestation and decline thereafter. Interestingly, a common feature among all the tissues that were chosen for investigating the regulation of T-18 expression, is the fact that they underwent apoptosis with increase in gestational age. The expression of T-18, in tissues exhibiting increased incidence of apoptosis suggested that T-18 maybe an apoptosis-associated gene. Using an explant culture model it was demonstrated that placental villi when cultured in vitro underwent spontaneous apoptosis and that the levels of T-18 message increased, under these conditions. Furthermore, this spontaneous induction of apoptosis in explant cultures could be blocked when villi were cultured in the presence of superoxide dismutase, a free radical scavenging enzyme. In addition, the expression of T-18 was shown to be modulated following treatment with SOD, or in response to oxidative stress. These studies clearly indicate a role for T-18 in placental apoptosis and moreover, implicate the usefulness of explant culture to examine the molecular mechanisms involved in placental apoptosis. Furthermore, the expression of the anti- and pro-apoptotic genes, bcl-x and bax respectively, were investigated, in an attempt to elucidate the signalling pathway(s) that led to the activation of an important downstream protease, caspase-3, in placental apoptosis. The present study revealed that induction of apoptosis in the placenta in vitro involved a bcl/bax independent, caspase-3 dependant pathway. The validation of an explant culture model for studying placental apoptosis and data pertaining to the role of T-18, bcl-x, bax and CPP32 in placental apoptosis, in response to oxidative stress, are presented in chapter 5. The last section titled general discussion summarizes the work carried out in this study and proposes a model for the apoptotic mechanism(s) that may be operating in placenta In conclusion, the present study has led to the identification of two developmentally-regulated factors, T-Plastin and T-18 in the first trimester and term placenta, respectively. The differential expression of these genes, in addition to several other molecular players, is proposed to be responsible for the overall functional differentiation of the placenta through the course of gestation.

Philosophiae Doctor - PhD / Drought is the most complex phenomenon that remained to be a potential and historic challenge to human welfare. It affects plant productivity by eliciting perturbations related to a pathway that controls a normal, functionally intact biological process of the plant. Sorghum (Sorghum bicolor (L.) Moench), a drought adapted model cereal grass is a potential target in the modem agricultural research towards understanding the molecular and cellular basis of drought tolerance. This study reports on the genomic and proteomic findings of drought tolerance in sorghum combining the results from in silica and experimental analysis. Pipeline that includes mapping expression data from 92 normalized cDNAs to genomic loci were used to identify drought tolerant genes. Integrative analysis was carried out using sequence similarity search, metabolic pathway, gene expression profiling and orthology relation to investigate genes of interest. Gene structure prediction was conducted using combination of ab initio and extrinsic evidence-driven information employing multi-criteria sources to improve accuracy. Gene ontology was used to cross-validate and to functionally assign and enrich genes. An integrated approach that subtly combines functional ontology based semantic data with expression profiling and biological networks was employed to analyse gene association with plant phenotypes and to identify and genetically dissect complex drought tolerance in sorghum. The gramene database was used to identify genes with direct or indirect association to drought related ontology terms in sorghum. Where direct association for sorghum genes were not available, genes were captured using Ensemble Biomart by transitive association based on the putative functions of sorghum orthologs in closely related species. Ontology mapping represented a direct or transitive association of genes to multiple drought related ontology terms based on sorghum specific genes or orthologs in related species. Correlation of genes to enriched gene ontology (GO)-terms (p-value < 0.05) related to the whole-plant structure was used to determine the extent of gene-phynotype association across-species and environmental stresses.