No description available.

This thesis proposes a novel recommendation approach to take advantage of the information available in user-created lists. Our approach assumes associations among any two items appearing in a list together. We consider two different ways to calculate the strength of item-item associations: frequency of co-occurrence, and sum of Bayesian ratings (SBR) of all lists containing the item pair. The latter takes into consideration not only the number of lists the items have co-appeared in, but also the quality of the lists. We collected a data set of user ratings for books along with Listmania lists on Amazon.com using Amazon Web Services (AWS). Our method shows superior performance to existing user-based and item-based collaborative filtering approaches according to the resulted Mean Absolute Error (MAE), coverage, precision and recall.

Submitted by Camila Urbano Braz null (camila_urbano@yahoo.com.br) on 2016-08-29T20:11:34Z No. of bitstreams: 1 Tese_Camila_Urbano_Braz.pdf: 2227912 bytes, checksum: 63cf8f89ad9895f8c74ed5d890e0b2f9 (MD5) / Approved for entry into archive by Juliano Benedito Ferreira (julianoferreira@reitoria.unesp.br) on 2016-08-30T18:25:24Z (GMT) No. of bitstreams: 1 braz_cu_dr_jabo.pdf: 2227912 bytes, checksum: 63cf8f89ad9895f8c74ed5d890e0b2f9 (MD5) / Made available in DSpace on 2016-08-30T18:25:24Z (GMT). No. of bitstreams: 1 braz_cu_dr_jabo.pdf: 2227912 bytes, checksum: 63cf8f89ad9895f8c74ed5d890e0b2f9 (MD5) Previous issue date: 2016-07-29 / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / O mercado consumidor tem sido cada vez mais exigente quanto à qualidade da carne e, portanto, a pecuária precisa melhorar sua eficiência e fornecer produtos diferenciados, padronizados e com qualidade. Entre as características de qualidade de carne, a maciez é a que mais influencia na satisfação dos consumidores. Considerando a importância dada à maciez da carne, pesquisas têm sido desenvolvidas para melhor compreender os mecanismos relacionados à expressão fenotípica desta característica. Os estudos de genes candidatos têm possibilitado a identificação de polimorfismos que modificam as estruturas das proteínas ou ainda, que estejam em desequilíbrio de ligação com alterações funcionais no DNA. Com a automatização dos polimorfismos de nucleotídeo único (SNPs) muitas regiões ao longo do genoma foram identificadas como responsáveis pela variação fenotípica da maciez da carne. No entanto, os marcadores SNPs podem ter baixa capacidade de identificar locos que atuam nas características quantitativas (QTL) por serem, na grande maioria, bi-alélicos e, mesmo que aconteçam mutações, as mudanças nas frequências alélicas dos SNPs podem permanecer quase inalteradas. Por outro lado, com a utilização de haplótipos, mutações tendem a causar mudanças nas frequências dos haplótipos, aumentando as chances de identificação dos QTL. Sendo assim, este estudo teve como objetivos detectar QTL e mutações causais em genes candidatos e identificar QTL por meio de uma análise de associação genômica ampla, para a característica maciez da carne em bovinos da raça Nelore, utilizando haplótipos como unidades fundamentais dos testes de associação. Foram utilizadas informações fenotípicas, genotípicas e de pedigree de animais provenientes de fazendas que integram os programas de melhoramento genético DeltaGen e PAINT. Cinquenta e dois genes candidatos foram escolhidos para serem analisados utilizando haplótipos construídos com base no desequilíbrio de ligação, utilizando 1.616 animais. Do total de haplótipos, dois foram significativos, sendo que os éxons próximos e dentro destes foram sequenciados visando buscar a mutação causal. O sequenciamento foi realizado com 298 animais e os SNPs identificados foram imputados para os 1.318 animais remanescentes. Foram realizadas análises de associação utilizando haplótipos construídos com base na metodologia de janelas sobrepostas, sendo que seus efeitos foram estimados pelo método Genomic Best Linear Unbiased Predictor (GBLUP). Os valores genéticos dos animais foram estimados para cada haplótipo e SNP e, após, as variâncias genéticas aditivas foram calculadas. Utilizando haplótipos construídos com base em janelas sobrepostas, verificou-se que o aumento do número de SNPs no haplótipo permitiu capturar maior proporção da variância genética aditiva da característica maciez da carne. Seis possíveis QTL foram identificados explicando as maiores proporções de variância genética aditiva para maciez da carne, dos quais um está no gene CAPN1 e cinco no gene ASAP1. Não houve evidências de que a mutação causal para a maciez da carne tenha sido identificada nos dois genes. Uma análise de associação genômica ampla foi realizada utilizando haplótipos construídos com base na metodologia de janelas sobrepostas de tamanhos variados. Foram utilizados nesta análise 1.405 animais genotipados com o painel Illumina Bovine HD e 1.756 animais genotipados com painel de menor densidade (70 K Neogen) e, posteriormente, imputados para o painel HD, em um total de 3.161 animais analisados. Os efeitos dos haplótipos e SNPs foram estimados pelo método GBLUP e as variâncias genética aditivas de cada haplótipo e SNP foram calculadas. Os genes NOS1AP, SUCLG1, PHLDB2 e LOC107132946 foram associados com a característica maciez da carne em bovinos da raça Nelore, por meio de análises de associação genômica ampla utilizando SNPs individuais e haplótipos. A análise utilizando SNPs identificou QTL diferentes das análises com haplótipos e, em alguns casos, SNPs apresentaram variâncias genéticas aditivas maiores do que as apresentadas pelos haplótipos. A análise que utilizou haplótipos construídos com cinco SNPs identificou mais QTL do que as análises de haplótipos construídos com sete e nove SNPs. Sugere-se que análises utilizando haplótipos, baseados em janelas sobrepostas, sejam realizadas para complementar análises de SNPs individuais em estudos de associação genômica ampla. / The consumer market has been increasingly demanding about the meat quality and therefore livestock needs to improve its efficiency and provide differentiated products, standardized and with quality. Considering the importance given to the meat tenderness, research has been undertaken to better understand the mechanisms related to the phenotypic expression of this trait. The candidate gene studies have allowed the identification of polymorphisms that change the structures of the proteins or that are in linkage disequilibrium with functional alterations in the DNA. With the advent of single nucleotide polymorphisms (SNPs) throughout many regions of the genome have been identified as responsible for phenotypic variation in meat tenderness. However, SNPs markers may have low ability to identify mutations, because SNPs are commonly bi-allelic and even when mutations have occurred, allelic frequencies can remain unaltered. On the other hand, using haplotype, mutations tend to cause major changes in haplotype frequencies, increasing the chances of identification of QTL. Thus, this study aimed to detect QTL and causal mutations by the approach of candidate genes and identify possible QTL through a genome-wide association analysis, for the trait meat tenderness in Nelore cattle using haplotypes as fundamental units of association tests. Information of the phenotypic, genotypic and pedigree were used from farms that belong to the breeding programs DeltaGen and PAINT. Fifty-two candidate genes were chosen for analysis using haplotypes constructed based on linkage disequilibrium using 1,616 animals. Two haplotypes were significant, and the exons near and within these haplotypes were sequenced to search for the causal mutation. The sequencing was performed using 298 animals and the identified SNPs were imputed for 1,318 remaining animals. Association analysis using haplotypes constructed based on the method of overlapping sliding windows were carried out and the SNPs and haplotypes effects were estimated using Genomic Best Linear Unbiased Predictor (GBLUP) method. The breeding values were estimated for each haplotype and SNPs and the additive genetic variances were calculated. Using haplotypes constructed based on overlapping sliding windows, we found that increasing the number of SNPs in the haplotype allowed to capture a greater proportion of additive genetic variance of meat tenderness. Six putative QTL were identified with the greatest additive genetic variances for meat tenderness, which one was in CAPN1 gene and five in ASAP1 gene. There was no evidence that the causal mutation for meat tenderness trait has been identified in these genes. A genome-wide association analysis was performed using haplotypes constructed based on the methodology of overlapping sliding windows of varying sizes. In this analysis, 1,405 animals genotyped with the Illumina Bovine HD panel and 1,756 genotyped animals with lower density panel (Neogen 70 K) were used and then, the genotypes of the 1,756 were imputed to the HD panel, in a total of 3,161 animals analyzed. The haplotypes and SNPs effects were estimated by the method GBLUP and the additive genetic variances were calculated for each haplotype and SNP. The NOS1AP, SUCLG1, PHLDB2 and LOC107132946 genes were associated with the meat tenderness trait in Nelore cattle through genome-wide association analysis using individual SNPs and haplotypes. The analysis using SNPs identified different QTL of the haplotype analyzes, and in some cases, the SNPs showed additive genetic variance greater than those presented by the haplotypes. The analysis used haplotypes constructed with five SNPs identified more QTL than analysis of haplotypes constructed with seven and nine SNPs. Analyzes using haplotypes based on overlapping sliding windows, should be conducted as additional analyzes for individual SNPs in genome-wide association studies. / FAPESP: 2013/00035-9

One of the major challenges for modern technology companies is user retentionmanagement. This work focuses on identifying early usage patterns that signifyincreased retention rates in a mobile web browser.This is done using a targetedparallel implementation of the association rule mining algorithm FP-Growth.Different item subset selection techniques including clustering and otherstatistical methods have been used in order to reduce the mining time and allowfor lower support thresholds.A lot of interesting rules have been mined. The best retention-wise ruleimplies a retention rate of 99.5%. The majority of the rules analyzed in thiswork implies a retention rate increase between 150% and 200%.

No description available.

Indiana University-Purdue University Indianapolis (IUPUI) / New neurons are generated throughout adulthood in two regions of the brain, the dentate gyrus of the hippocampus, which is important for memory formation and cognitive functions, and the sub-ventricular zone of the olfactory bulb, which is important for the sense of smell, and are incorporated into hippocampal network circuitry. Disruption of this process has been postulated to contribute to neurodegenerative disorders including Alzheimer’s disease [1]. AD is the most common form of adult-onset dementia and the number of patients with AD escalates dramatically each year. The generation of new neurons in the dentate gyrus declines with age and in AD. Many of the molecular players in AD are also modulators of adult neurogenesis, but the genetic mechanisms influencing adult neurogenesis in AD are unclear. The overall goal of this project is to identify candidate genes and pathways that play a role in neurogenesis in the adult brain and to test the hypotheses that 1) hippocampal neurogenesis-related genes and pathways are significantly perturbed in AD and 2) neurogenesis-related pathways are significantly associated with hippocampal volume and other AD-related biomarker endophenotypes including brain deposition of amyloid-β and tau pathology. First, potential modulators of adult neurogenesis and their roles in neurodegenerative diseases were evaluated. Candidate genes that control the turnover process of neural stem cells/precursors to new functional neurons during adult neurogenesis were manually curated using a pathway-based systems biology approach. Second, a targeted neurogenesis pathway-based gene analysis was performed resulting in the identification of ADORA2A as associated with hippocampal volume and memory performance in mild cognitive impairment and AD. Third, a genome-wide gene-set enrichment analysis was conducted to discover associations between hippocampal volume and AD related endophenotypes and neurogenesis-related pathways. Within the discovered neurogenesis enriched pathways, a gene-based association analysis identified TESC and ACVR1 as significantly associated with hippocampal volume and APOE and PVLR2 as significantly associated with tau and amyloid beta levels in cerebrospinal fluid. This project identifies new genetic contributions to hippocampal neurogenesis with translational implications for novel therapeutic targets related to learning and memory and neuroprotection in AD.

No description available.

<p>Complex traits are influenced by genetic and environmental factors and their interactions. Most common human disorders such as cardiovascular, metabolic, autoimmune, and neurological diseases are complex. Understanding their genetic architecture and etiology is an important step to prevent, diagnose and treat these conditions. Genome Wide Association Studies (GWAS) have emerged as a powerful and widely used tool that can be used to explore and identify the genetic variants associated with complex traits. In this dissertation, we present some of the downstream applications of GWAS studies to analyze and understand the genetic risk and etiology of complex traits and provide important insights into the genetic architecture and background of several complex phenotypes. First, we examined whether prevalence of complex disorders around the world correlates to Polygenic Risk Scores (PRS). To do so, we determined the average PRS of 14 such complex disorders across 24 world populations using results of GWAS studies. We found variation in risk across populations and significant correlation was obtained between average disease risk and prevalence for seven of the studied disorders. Further exploring the power of PRS- based calculations, we performed a PRS - based phenome wide association study (PheWAS) for Tourette Syndrome (TS) and identified 57 phenotypic outcomes significantly associated with TS PRS. The strongest associations were found between TS PRS and mental health factors. Cross- disorder comparisons of phenotypic associations with genetic risk for other childhood-onset disorders (e.g.: attention deficit hyperactivity disorder [ADHD], autism spectrum disorder [ASD], and obsessive-compulsive disorder [OCD]) indicated an overlap in associations between TS and these disorders. Furthermore, we performed a sex specific PheWAS that highlighted differences in associations of complex disorders with TS PRS in males and females. Finally, we used large- scale GWAS results to identify causal associations between different biological markers (proteins, metabolites, and microbes) and subcortical brain structure volumes using Mendelian Randomization (MR) analysis. We identified eleven proteins and six metabolites to be significantly associated with subcortical brain volume structures. Enrichment analysis indicated that the associated proteins were enriched for proteolytic functions and regulation of apoptotic pathways. Overall, our work demonstrates the power of GWAS studies to help disentangle the genetic basis of complex diseases and also provides important insights into the etiology of the studied complex traits. </p>

No description available.

Les patrons de variation au sein des espèces d'arbres résultent d'évolutions complexes dont certaines sont particulièrement liées à la condition d'arbre. Si les facteurs expliquant cette variabilité ont été abordés pour les espèces des zones tempérées, ils restent encore peu connus pour des espèces fruitières des zones tropicales sèches offrant des particularités sur le plan adaptatif et relevant d'un processus de domestication ancien. Pour répondre à cet enjeu, l'objectif scientifique de cette thèse est de mieux comprendre l'impact des facteurs évolutifs sur la variabilité au sein d'une espèce d'arbre en zone soudano-sahélienne. Nous nous sommes intéressés à Vitellaria paradoxa (karité), une espèce jouant un rôle économique majeur par la production de beurre à partir des graines, dont l'aire de distribution s'étend du Sénégal à l'Ouganda sous la forme de peuplements naturels et agroforestiers. Dans une approche phylogéographique basée sur l'analyse du polymorphisme de marqueurs moléculaires supposés neutres (microsatellites chloroplastiques et nucléaires, et régions intergéniques du chloroplaste) et sur la modélisation des niches écologiques du Karité à différentes périodes du Quaternaire, nous mettons en évidence l'impact des perturbations liées aux dernières glaciations (il y a 20000 ans) sur la diversité génétique de l'espèce. Dans une seconde approche méthodologique, nous étudions l'impact des déterminants environnementaux sur la variabilité des constituants chimiques des graines de Karité, permettant d'énoncer de nouvelles hypothèses. Enfin, nous nous intéressons à l'expression du polymorphisme de gènes candidats, codant les enzymes stéaroyl-ACP-désaturase (SAD) et homogentisate phytyltransférase (HPT), en relation avec la variabilité de la composition chimique en acides gras et en tocophérols des graines. Les résultats obtenus permettent notamment de discuter de l'impact de la sélection naturelle, de la domestication et de la dérive génétique, et d'apporter des hypothèses évolutives et fonctionnelles potentielles expliquant la variabilité observée. / Patterns of variation within tree species are the result of complex evolutionary changes, some of which are particularly related to the condition of trees. If the factors explaining this variability were discussed for species in temperate zones, they remain little known for fruit tree species of dry tropics, which show special adaptive features and result from an old domestication process. To meet this challenge, the scientific purpose of this thesis is to improve our understanding of evolutionary factors affecting the variability within a tree species in the Sudano-Sahelian zone. We got interested in Vitellaria paradoxa (Shea tree), a species which plays a major economic role for the production of butter from seeds and whose distribution range extends from Senegal to Uganda. In a phylogeographic approach based on the analysis of polymorphism supposedly neutral molecular markers (chloroplast and nuclear microsatellites and chloroplast intergenic regions) and based on ecological niche modelling of Shea tree at different periods of the Quaternary, we shape the role of last glacial maximum (20.000 YBP) on the genetic diversity of the species. In a second methodological approach, we study the impact of environmental determinants on the variability of Shea nuts chemical constituents, stating novel hypotheses. Finally, we investigate the polymorphism of candidate genes, encoding enzymes stearoyl-ACP-desaturase (SAD) and homogentisate phytyltransférase (HPT), in connection with the variability of relative fatty acids compositions and tocopherols contents in seeds. Through the results obtained, we discuss the impact of natural selection, domestication and genetic drift, and provide evolutionary and functional hypotheses that potentially explain the varia bility observed.