The evolutionary relationship between Salmonella enterica serovar Typhi, other typhoid-like enteric fever causing serovars and 10 non-Typhoid serovars from S. enterica subspecies I, could not be determined by comparative nucleotide sequences of six genes. Phylogenetic analyses of the dataset showed that the genes of interest underwent frequent recombination, suggesting a low level of clonality within subspecies I of S. enterica. To establish the evolutionary relationships within serovar Typhi, genome-wide Single Nucleotide Polymorphism (SNP) was explored as a marker for both typing purposes and phylogenetic analysis. Thirty eight SNPs were typed in 73 global Typhi isolates, including 18 isolates expressing the special flagellar antigen z66, using restriction enzyme digestion method. The isolates were differentiated into 23 SNP profiles and grouped into four distinct clusters. The z66 isolates were divided into four SNP profiles and were all grouped into one cluster, suggesting a single origin. An alternative SNP typing method using the hairpin real time PCR assay was investigated to type four additional SNPs, termed as biallelic polymorphisms (BiP). These BiPs were found to classify 481 global Typhi isolates into five major clusters (Roumagnac et al., 2006). Typing four BiPs resulted in the identification of four additional SNP profiles. We proposed nine SNPs were required to type Typhi isolates into 13 subclusters for global epidemiology. An enzymatic-based method using CelI nuclease was evaluated to discover more SNPs from other Typhi genomes. The efficiency of the CelI was shown to be unsatisfactory and we were unable to demonstrate the effectiveness of the proposed method. Nine Variable Number of Tandem Repeats (VNTRs) were typed in the 73 Typhi isolates using fluorescent-labelled universal primers, and analysed on an automated DNA sequencer. Five isolates were unable to give PCR products in one or more VNTR loci. Nine VNTRs could differentiate 68 Typhi isolates into 65 MLVA profiles, suggesting a higher discriminating power than SNP typing. SNPs were shown to be a more appropriate marker for phylogenetic tracing for Typhi while VNTRs were highly discriminating but could not be used to establish the evolutionary relationships of diverse Typhi isolates.

Otitis Media is one of the most common childhood diseases. Recurrent acute otitis media RAOM is characterized by repeated episodes of inflammation of the middle ear in conjunction with middle ear fluid, and often with an inflamed or bulging eardrum. Defective clearance by the Eustachian tube results in mucus build-up and is characteristic of otitis media with effusion (OME). Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis, respiratory syncytial virus, and rhinovirus are the most common contributors to otitis media pathogenesis. In New Zealand, OME has been implicated with conductive hearing loss in childhood and has been shown to significantly impact on speech and language development. New Zealand Māori and Polynesian children have displayed significantly higher hearing test failure rates than European-Caucasian children. The collectins, Surfactant Protein (SP)-A and -D are encoded by three genes (SP-A1, SP-A2, and SP-D) and are host defense proteins present in the middle ear and Eustachian tube. Single nucleotide polymorphisms (SNPs) in SP-A1 and SP-A2 have been associated with increased or decreased susceptibility to otitis media, meningococcal disease, and range of respiratory diseases. Using allele-specific primers and real-time PCR with SYBR Green I melting curve analysis, four groups of individuals were genotyped for eleven SP-A1, SP-A2, and SP-D SNPs: European-Caucasian individuals with RAOM/OME; New Zealand Māori/Polynesian individuals with RAOM/OME; individuals with meningococcal disease; and a control group. The computer program, Haploview, was employed to perform χ2 analyses and identify statistically significant associations of alleles/haplotypes with RAOM/OME or meningococcal disease. In the European-Caucasian population, two SP-A1 alleles, one SP-A2 allele, and four haplotypes (CGAGC, 1A3, 1A9, and 1A10) were found to be associated with increased risk of RAOM/OME (P lt; 0.05). Conversely, haplotypes 6A2 and 1A2 were found to be protective against susceptibility to RAOM/OME (P lt; 0.05). In New Zealand Māori and Polynesian individuals, two SP-A1 alleles, three SP-A2 alleles, one SP-D allele, and four haplotypes (6A8, 6A10, 1A3, and 1A10) were found to be associated with increased risk of RAOM/OME (P lt; 0.05). An additional four haplotypes (6A2, 1A0, 1A2, and TA) were determined to be protective against susceptibility to RAOM/OME (P lt; 0.05). However, protective SPA1/SPA2/SPD haplotype 6A2-1A0-TA was significantly under-represented in the New Zealand Māori and Polynesian population (P lt; 0.05). A single allele and haplotype were associated with increased risk of meningococcal disease (P lt; 0.05). The findings of this study confirm that specific genetic variants of SP-A and SP-D are associated with either increased or decreased risk of developing RAOM and/or OME. Furthermore, it was demonstrated that New Zealand Māori and Polynesian individuals appear to exhibit more haplotypes susceptible to RAOM/OME. This may provide a partial explanation for the higher RAOM/OME-related failure rates of hearing tests in New Zealand Māori and Polynesian children. However, there are numerous socio-economic and environmental factors that also contribute to otitis media pathogenesis which were not considered in this study. The effects of the SP-A1, SP-A2, and SP-D alleles and haplotypes on the bacterial/viral binding efficiencies of SP-A and SP-D need to be investigated by further research, using a large population, to confirm the association with susceptibility or resistance with RAOM/OME.

Radiotherapy is often used to treat prostate cancer. While using high dose of radiation does kill cancer cells, it can cause toxicity in healthy tissues for some patients. It would be best to apply this treatment only to patients who are likely to be immune from such toxicity. This requires a classifier that can predict, before treatment, which patients are likely to exhibit severe toxicity. Here, we explore ways to use certain genetic features, called Single Nucleotide Polymorphisms (SNPs), for this task. This thesis uses several machine learning methods for learning such classifiers for predicting toxicity. This problem is challenging as there are a large number of features (164,273 SNPs) but only 82 samples. We explore an ensemble classification method for this problem, called Mixture Using Variance (MUV), which first learns several different base probabilistic classifiers, then for each query combines the responses of the different base classifiers based on their respective variances. The original MUV learns the individual classifiers using bootstrap sampling of the training data; we modify this by considering different subsets of the features for each classifier. We derive a new combination rule for base classifiers in the proposed setting and obtain some new theoretical results. Based on characteristics of our task, we propose an approach that involves first clustering the features before selecting only a subset of features from each cluster for each base classifier. Unfortunately, we were unable to predict radiation toxicity in prostate cancer patients using just the SNP values. However, our further experimental results reveal strong relation between correctness of a classifier in its prediction and the variance of the response to the corresponding classification query, which show that the main idea is promising.

Autism is a severe childhood neuropsychiatric condition characterized by impairments in socialization and communication, and by restricted and repetitive behaviours. Autism spectrum disorder (ASD) is a complex and largely unknown disease with a strong genetic basis, multiple genes involved and environmental factors determining its phenotype. Interestingly, the DLX1/DLX2 and DLX5/DLX6 bigene clusters are located in autism susceptibility loci and Dlx genes are involved in GABAergic interneurons differentiation and migration to the cortex during forebrain development. Dlx gene expression is controlled by different cis-regulatory elements. Of these, 4 are active in the forebrain, URE2, I12b, I56ii and I56i. In order to determine the role of the DLX genes in ASD, variants were found in gene exons and in cis-regulatory elements in autistic individuals. A single nucleotide polymorphism (SNP), a change of an adenine for a guanine, was identified in I56i enhancer. Finding a SNP in I56i was very surprising considering that it is located in a Dlx binding motif highly conserved among >40 species. We showed, using in vitro approaches, that the presence of this SNP affects the affinity of Dlx for their binding site and reduces the transcriptional activation of the enhancer. The SNP also affects activity of the I56i enhancer in transgenic mice. In order to determine the real impact of the SNP in vivo, mutant mice harboring the SNP in their I56i enhancer were produced. That involved the insertion of the I56i enhancer with the SNP, using homologous recombination in mouse embryonic stem cells to replace the wild type version of the enhancer. With these mutant mice, we demonstrated that, in vivo, this SNP reduces Dlx5 and Dlx6 expression in the forebrain. Furthermore, this decrease in Dlx5/Dlx6 expression could affect the differentiation and/or migration of specific populations of inhibitory interneurons in the forebrain. No distinct iv behavioural phenotypes were observed between wild type mice and those carrying the SNP, during social interaction and anxiety tests. Therefore, these results suggest that even a subtle change in a regulatory element can have an impact in the development of the forebrain and may even contribute to disorders such as autism.

The most intriguing problems in genetics epidemiology are to predict genetic disease susceptibility and to associate single nucleotide polymorphisms (SNPs) with diseases. In such these studies, it is necessary to resolve the ambiguities in genetic data. The primary obstacle for ambiguity resolution is that the physical methods for separating two haplotypes from an individual genotype (phasing) are too expensive. Although computational haplotype inference is a well-explored problem, high error rates continue to deteriorate association accuracy. Secondly, it is essential to use a small subset of informative SNPs (tag SNPs) accurately representing the rest of the SNPs (tagging). Tagging can achieve budget savings by genotyping only a limited number of SNPs and computationally inferring all other SNPs. Recent successes in high throughput genotyping technologies drastically increase the length of available SNP sequences. This elevates importance of informative SNP selection for compaction of huge genetic data in order to make feasible fine genotype analysis. Finally, even if complete and accurate data is available, it is unclear if common statistical methods can determine the susceptibility of complex diseases. The dissertation explores above computational problems with a variety of methods, including linear algebra, graph theory, linear programming, and greedy methods. The contributions include (1)significant speed-up of popular phasing tools without compromising their quality, (2)stat-of-the-art tagging tools applied to disease association, and (3)graph-based method for disease tagging and predicting disease susceptibility.

This study focuses how the MLR-tagging for statistical covering, i.e. either maximizing average R2 for certain number of requested tags or minimizing number of tags such that for any non-tag SNP there exists a highly correlated (squared correlation R2 > 0.8) tag SNP. We compare with tagger, a software for selecting tags in hapMap project. MLR-tagging needs less number of tags than tagger in all 6 cases of the given test sets except 2. Meanwhile, Biologists can detect or collect data only from a small set. So, this will bring a problem for scientists that the estimates accuracy of tag SNPs when constructing the complete human haplotype map. This study investigates how the MLR-tagging for statistically coverage performs under unbias study. The experiment results shows MLR-tagging still select small amount of SNPs very well even without observing the entire SNP in the sample.

Autism is a severe childhood neuropsychiatric condition characterized by impairments in socialization and communication, and by restricted and repetitive behaviours. Autism spectrum disorder (ASD) is a complex and largely unknown disease with a strong genetic basis, multiple genes involved and environmental factors determining its phenotype. Interestingly, the DLX1/DLX2 and DLX5/DLX6 bigene clusters are located in autism susceptibility loci and Dlx genes are involved in GABAergic interneurons differentiation and migration to the cortex during forebrain development. Dlx gene expression is controlled by different cis-regulatory elements. Of these, 4 are active in the forebrain, URE2, I12b, I56ii and I56i. In order to determine the role of the DLX genes in ASD, variants were found in gene exons and in cis-regulatory elements in autistic individuals. A single nucleotide polymorphism (SNP), a change of an adenine for a guanine, was identified in I56i enhancer. Finding a SNP in I56i was very surprising considering that it is located in a Dlx binding motif highly conserved among >40 species. We showed, using in vitro approaches, that the presence of this SNP affects the affinity of Dlx for their binding site and reduces the transcriptional activation of the enhancer. The SNP also affects activity of the I56i enhancer in transgenic mice. In order to determine the real impact of the SNP in vivo, mutant mice harboring the SNP in their I56i enhancer were produced. That involved the insertion of the I56i enhancer with the SNP, using homologous recombination in mouse embryonic stem cells to replace the wild type version of the enhancer. With these mutant mice, we demonstrated that, in vivo, this SNP reduces Dlx5 and Dlx6 expression in the forebrain. Furthermore, this decrease in Dlx5/Dlx6 expression could affect the differentiation and/or migration of specific populations of inhibitory interneurons in the forebrain. No distinct iv behavioural phenotypes were observed between wild type mice and those carrying the SNP, during social interaction and anxiety tests. Therefore, these results suggest that even a subtle change in a regulatory element can have an impact in the development of the forebrain and may even contribute to disorders such as autism.

No description available.

No description available.

No description available.