The 1941 Binning House in West Vancouver has long been hailed as a pioneer of Modernism in domestic Canadian architecture, and an inspiration for much of the West Coast Architecture that followed. Although it is usually described as product of Corbusian rationalism and a paradigm of low-cost dwelling, in fact it is neither. Rather, it is a composite of several competing strains of Modernism and aesthetic values prevalent in London during the year (1938-39) in which Binning resided there to study fine art. The Binning House is often misread as an austerely functionalist plan with an orthogonal layout, but a closer observation and actual measurement of wall and window angles reveals that Binning actually inflected the orthogonal, generating a splayed geometric layout with obtuse and acute angles in several corners, trapezoidal forms in the built-in furniture and studio clerestory window, and a dynamic sense of visual expansion and contraction. Binning's study with Henry Moore was evidently tremendously influential in this regard, as Moore avoided the machine-like aesthetic of the orthogonal and instead imbued his art with oblique, irregular and rounded lines. The oblique motif also manifests in Binning's own drawings of this time. Also empathetic to this approach was Berthold Lubetkin, whose Whipsnade Bungalow near London defied the doctrines of orthogonal functionalism. Binning viewed plans and photos of Whipsnade and other emblems of early European modernism at a seminal 1939 exhibition at the Museum of Modern Art in New York. This exhibition synthesized many of the ideas and forms that Binning had been exposed to in London and seems to have served as a catalyst for the house plan he was about to compose. In converging these various strains of early Modernism, Binning has transcended the dogma of architectural discourse and rendered it meaningful for a local, individual context. / Applied Science, Faculty of / Architecture and Landscape Architecture (SALA), School of / Graduate

Bertram Charles Binning's depiction of British Columbia coastal scenes in his drawings and paintings of the mid 1940s to the early 1950s present images of sunlit seascapes in recreational settings; they are scenes of leisure and pleasure. The concern for leisure and pleasure was central to the artist's modernism, even after he began painting in a semi-abstract manner around 1948. In this particular construction of modernism, Binning offered pleasure as an antidote to some of the anxieties he observed in postwar culture. Binning also thought that art could contribute to life in a direct way. In the mid to late 1940s, Vancouver saw a series of artistic community projects which explored the possibility of art as a social force; the Art in Living Group, of which Binning was a member, believed that art could have a therapeutic value in relation to housing projects and community planning. In certain ways, the Art in Living Group was a response to rapid changes in the social matrix of Vancouver. Binning's personal artistic practice, however, appears to have existed outside of what was embraced in his participation in those community projects. His essentially personal, self-authenticating expression in the form of drawings may be seen to resist the idealism of his more 'public' production, that is, his own idealism, his demand for an art thoroughly harmonized with the public sphere. Moreover, in this more personal body of work, his choice of leisurely scenes, rendered in a style reminiscent of Matisse, can be seen as far removed from the urban tensions of the time. It also seems to suggest that the leisure-and-pleasure idealism which finds expression in these works was not only class-and gender- specific, but also antithetical to his strong desire to democratize art. Binning's preoccupation with personal expression took a turn when he shifted his concern from representational drawings to semi-abstract paintings. The shift coincided with his career move to the University of British Columbia as a professor of Art History in 1949. From then on, Binning's interest in regional cosmopolitanism became more pronounced in his work. In this sense, it is significant that Binning looked for guidance to Herbert Read's ideas about modern art and art education. At the same time, his reputation expanded beyond the West Coast. In 1954, Binning was chosen to represent Canada at the Venice Biennale. Binning's particular modernism, as represented by this range of work, all of which presents a pastoral version of Utopia , was in some ways profoundly at odds with the social circumstances of the time. Why was the interest in leisure and pleasure significant to his practice? What did it mean to promote this kind of idealism in the local context? And in what ways did it relate to the international art scene — for example, to the work of Matisse or to contemporary concepts of art? My thesis addresses these questions by situating Binning's work both regionally and internationally.

Bertram Charles Binning's depiction of British Columbia coastal scenes in his drawings and paintings of the mid 1940s to the early 1950s present images of sunlit seascapes in recreational settings; they are scenes of leisure and pleasure. The concern for leisure and pleasure was central to the artist's modernism, even after he began painting in a semi-abstract manner around 1948. In this particular construction of modernism, Binning offered pleasure as an antidote to some of the anxieties he observed in postwar culture. Binning also thought that art could contribute to life in a direct way. In the mid to late 1940s, Vancouver saw a series of artistic community projects which explored the possibility of art as a social force; the Art in Living Group, of which Binning was a member, believed that art could have a therapeutic value in relation to housing projects and community planning. In certain ways, the Art in Living Group was a response to rapid changes in the social matrix of Vancouver. Binning's personal artistic practice, however, appears to have existed outside of what was embraced in his participation in those community projects. His essentially personal, self-authenticating expression in the form of drawings may be seen to resist the idealism of his more 'public' production, that is, his own idealism, his demand for an art thoroughly harmonized with the public sphere. Moreover, in this more personal body of work, his choice of leisurely scenes, rendered in a style reminiscent of Matisse, can be seen as far removed from the urban tensions of the time. It also seems to suggest that the leisure-and-pleasure idealism which finds expression in these works was not only class-and gender- specific, but also antithetical to his strong desire to democratize art. Binning's preoccupation with personal expression took a turn when he shifted his concern from representational drawings to semi-abstract paintings. The shift coincided with his career move to the University of British Columbia as a professor of Art History in 1949. From then on, Binning's interest in regional cosmopolitanism became more pronounced in his work. In this sense, it is significant that Binning looked for guidance to Herbert Read's ideas about modern art and art education. At the same time, his reputation expanded beyond the West Coast. In 1954, Binning was chosen to represent Canada at the Venice Biennale. Binning's particular modernism, as represented by this range of work, all of which presents a pastoral version of Utopia , was in some ways profoundly at odds with the social circumstances of the time. Why was the interest in leisure and pleasure significant to his practice? What did it mean to promote this kind of idealism in the local context? And in what ways did it relate to the international art scene — for example, to the work of Matisse or to contemporary concepts of art? My thesis addresses these questions by situating Binning's work both regionally and internationally. / Arts, Faculty of / Art History, Visual Art and Theory, Department of / Graduate

No description available.

The last decade has seen great advances in Next-Generation Sequencing technologies, and, as a result, there has been a rise in the number of genomes sequenced each year. In 2017, there were as many as 10,000 new organisms sequenced and added into the RefSeq Database. Taxonomy prediction is a science involving the hierarchical classification of DNA fragments up to the rank species. In this research, we introduce Predicting Linked Organisms, Plinko, for short. Plinko is a fully-functioning, state-of-the-art predictive system that accurately captures DNA - Taxonomy relationships where other state-of-the-art algorithms falter. Plinko leverages multi-view convolutional neural networks and the pre-defined taxonomy tree structure to improve multi-level taxonomy prediction. In the Plinko strategy, each network takes advantage of different word usage patterns corresponding to different levels of evolutionary divergence. Plinko has the advantages of relatively low storage, GPGPU parallel training and inference, making the solution portable, and scalable with anticipated genome database growth. To the best of our knowledge, Plinko is the first to use multi-view convolutional neural networks as the core algorithm in a compositional,alignment-free approach to taxonomy prediction. / Master of Science / Taxonomy prediction is a science involving the hierarchical classification of DNA fragments up to the rank species. Given species diversity on Earth, taxonomy prediction gets challenging with (i) increasing number of species (labels) to classify and (ii) decreasing input (DNA) size. In this research, we introduce Predicting Linked Organisms, Plinko, for short. Plinko is a fully-functioning, state-of-the-art predictive system that accurately captures DNA - Taxonomy relationships where other state-of-the-art algorithms falter. Three major challenges in taxonomy prediction are (i) large dataset sizes (order of 109 sequences) (ii) large label spaces (order of 103 labels) and (iii) low resolution inputs (100 base pairs or less). Plinko leverages multi-view convolutional neural networks and the pre-defined taxonomy tree structure to improve multi-level taxonomy prediction for hard to classify sequences under the three conditions stated above. Plinko has the advantage of relatively low storage footprint, making the solution portable, and scalable with anticipated genome database growth. To the best of our knowledge, Plinko is the first to use multi-view convolutional neural networks as the core algorithm in a compositional, alignment-free approach to taxonomy prediction.

L’avènement du séquençage métagénomique aléatoire a révolutionné la microbiologie en permettant la caractérisation sans culture préalable de communautés microbiennes complexes telles que le microbiote intestinal humain. Des outils bioinformatiques récemment développés atteignent une résolution au niveau de la souche en recensant des gènes accessoires ou en capturant des variants nucléotidiques (SNPs). Toutefois, ces outils sont limités par l’étendue des génomes de référence disponibles qui sont loin de couvrir toute la variabilité microbienne. En effet, de nombreuses espèces n’ont pas encore été séquencées ou sont représentées par seulement quelques génomes.La création de catalogues de gènes non redondants par assemblage de novo suivie du regroupement des gènes co-abondants révèlent une partie de la matière noire microbienne en reconstituant le répertoire de gènes d’espèces potentiellement inconnues. Bien que les méthodes existantes identifient avec précision les gènes core présents dans toutes les souches d’une espèce, elles omettent de nombreux gènes accessoires ou les divisent en petits groupes de gènes qui ne sont pas associés aux core génomes. Or, capturer ces gènes accessoires est indispensable en recherche clinique et épidémiologique car ces derniers assurent des fonctions spécifiques à certaines souches telles que la pathogénicité ou la résistance aux antibiotiques.Lors de cette thèse, nous avons développé MSPminer, un logiciel performant qui reconstitue et structure des pan-génomes d’espèces métagénomiques (ou MSPs pour Metagenomic Species Pan-genomes) en regroupant les gènes co-abondants dans un ensemble d’échantillons métagénomiques. MSPminer s’appuie sur une nouvelle mesure robuste de la proportionnalité couplée à un classificateur empirique pour regrouper et distinguer les gènes core mais aussi les gènes accessoires des espèces microbiennes.Grâce à MSPminer, nous avons structuré un catalogue de 9,9 millions de gènes du microbiote intestinal humain en 1 661 MSPs. L’homogénéité de l’annotation taxonomique, de la composition nucléotidique ainsi que la présence de gènes essentiels indiquent que les MSPs ne correspondent pas à des chimères mais à des objets biologiquement cohérents regroupant des gènes provenant de la même espèce. Parmi ces MSPs, 1 301 (78%) n’ont pas pu être annotées au niveau espèce montrant que de nombreux microorganismes colonisant l’intestin humain demeurent inconnus malgré les progrès substantiels des techniques de culture microbienne. Remarquablement, les MSPs capturent bien plus de gènes que les clusters générés par les outils existants tout en garantissant une spécificité élevée.Cet ensemble de MSPs peut d’ores et déjà être utilisé pour le profilage taxonomique et la découverte de biomarqueurs dans des échantillons de selles humaines. Ainsi, nous tirons parti des MSPs pour comparer l’impact sur le microbiote intestinal des deux principaux types de chirurgie bariatrique, la gastrectomie par laparoscopie (LSG) et la dérivation gastrique de Roux-en-Y (LRYGB). Enfin, les MSPs ouvrent la voie à des analyses au niveau souche. Dans une autre cohorte, nous avons mis en évidence l’existence de sous-espèces associées à l’origine géographique de l’hôte en étudiant les profils de présence/absence des gènes accessoires groupés dans les MSPs. / The advent of shotgun metagenomic sequencing has revolutionized microbiology by allowing culture-independent characterization of complex microbial communities such as the human gut microbiota. Recently developed bioinformatics tools achieved strain-level resolution by making a census of accessory genes or by capturing nucleotide variants (SNPs). Yet, these tools are hampered by the extent of available reference genomes which are far from covering all the microbial variability. Indeed, many species are still not sequenced or are represented by only few genomes.Building of non-redundant gene catalogs followed by the binning of co-abundant genes reveals a part of the microbial dark matter by reconstituting the gene repertoire of species potentially unknown. While existing methods accurately identify core genes present in all the strains of a species, they miss many accessory genes or split them into small gene groups that remain unassociated to core genomes. However, capturing these accessory genes is essential in clinical research and epidemiology because they provide functions specific to certain strains such as pathogenicity or antibiotic resistance.In this thesis, we developed MSPminer, a computationally efficient software tool that reconstitutes Metagenomic Species Pan-genomes (MSPs) by binning co-abundant genes across metagenomic samples. MSPminer relies on a new robust measure of proportionality coupled with an empirical classifier to group and distinguish not only species core genes but accessory genes also.With MSPminer, we structured a catalog made up of 9.9 million genes of the human gut microbiota in 1 661 MSPs. The homogeneity of the taxonomic annotation, of the nucleotide composition as well as the presence of essential genes indicate that the MSPs do not correspond to chimeras but to biologically consistent objects grouping genes from the same species. Among these MSPs, 1 301 (78%) could not be annotated at species level showing that many microorganisms colonizing the human intestinal tract are still unknown despite the substantial improvements of microbial culture techniques. Remarkably, MSPs capture more genes than clusters generated by existing tools while ensuring high specificity.This set of MSPs can be readily used for taxonomic profiling and biomarkers discovery in human gut metagenomic samples. In this way, we take advantage of the MSPs to compare the impact of two main types of surgeries, the laparoscopic sleeve gastrectomy (LSG) and the Roux-En-Y gastric bypass (LRYGB). Finally, the MSPs open the way to strain-level analyses. In another cohort, we identified subspecies associated the host geographical origin by studying presence/absence patterns of the accessory genes grouped in the MSPs.

Although deep neural networks (DNNs) have made remarkable achievementsin various elds, there is still not a matching practical theory that is able toexplain DNNs' performances. Tishby (2015) proposed a new insight to analyzeDNN via the Information bottleneck (IB) method. By visualizing how muchrelevant information each layer contains in input and output, he claimed thatthe DNNs training is composed of tting phase and compression phase. Thetting phase is when DNNs learn information both in input and output, andthe prediction accuracy goes high during this process. Afterwards, it is thecompression phase when information in output is preserved while unrelatedinformation in input is thrown away in hidden layers. This is a tradeo betweenthe network complexity (complicated DNNs lose less information in input) andprediction accuracy, which is the same goal with the IB method.In this thesis, we verify this IB interpretation rst by reimplementing Tishby'swork, where the hidden layer distribution is approximated by the histogram(binning). Additionally, we introduce various mutual information estimationmethods like kernel density estimators. Based upon simulation results, we concludethat there exists an optimal bound on the mutual information betweenhidden layers with input and output. But the compression mainly occurs whenthe activation function is \double saturated", like hyperbolic tangent function.Furthermore, we extend the work to the simulated wireless model where thedata set is generated by a wireless system simulator. The results reveal that theIB interpretation is true, but the binning is not a correct tool to approximatehidden layer distributions. The ndings of this thesis reect the informationvariations in each layer during the training, which might contribute to selectingtransmission parameter congurations in each frame in wireless communicationsystems. / Ä ven om djupa neuronnät (DNN) har gjort anmärkningsvärda framsteg på olikaområden, finns det fortfarande ingen matchande praktisk teori som kan förklara DNNs prestanda. Tishby (2015) föreslog en ny insikt att analysera DNN via informationsflaskhack (IB) -metoden. Genom att visualisera hur mycket relevant information varje lager innehåller i ingång och utgång, hävdade han att DNNs träning består av monteringsfas och kompressionsfas. Monteringsfasenär när DNN lär sig information både i ingång och utgång, och prediktionsnoggrannheten ökar under denna process. Efteråt är det kompressionsfasen när information i utgången bevaras medan orelaterad information i ingången kastas bort. Det här är en kompromiss mellan nätkomplexiteten (komplicerade DNN förlorar mindre information i inmatning) och predictionsnoggrannhet, vilket är exakt samma mål med informationsflaskhals (IB) -metoden.I detta examensarbete kontrollerar vi denna IB-framställning först genom att implementera om Tishby’s arbete, där den dolda lagerfördelningen approximeras av histogrammet (binning). Dessutom introducerar vi olika metoder förömsesidig information uppskattning som kernel density estimators. Baserat på simuleringsresultatet drar vi slutsatsen att det finns en optimal bindning för denömsesidiga informationen mellan dolda lager med ingång och utgång. Men komprimeringen sker huvudsakligen när aktiveringsfunktionen är “dubbelmättad”, som hyperbolisk tangentfunktion.Dessutom utvidgar vi arbetet till den simulerad trådlösa modellen där data set genereras av en trådlös systemsimulator. Resultaten visar att IB-framställning är sann, men binningen är inte ett korrekt verktyg för att approximera dolda lagerfördelningar. Resultatet av denna examensarbete reflekterar informationsvariationerna i varje lager, vilket kan bidra till att välja överföringspa-rameterns konfigurationer i varje ram i trådlösa kommunikationssystem

The development of scorecards for customer credit rating is a well-established field in the financial sector. The aim of this project, conducted in collaboration with a Swedish credit institute, was to develop a statistical model for predicting customer performance. In addition to conducting a model, the project also sought to identify the set of consumer characteristics with high predictive capability and how these characteristics differ when predicting performance early versus late in the loan term. To achieve this goal, a dataset containing approximately 15,000 unique loan applications approved between July 2020 and July 2022 was acquired from the credit institute, and logistic regression models were applied for different time periods ranging from 6 to 21 months. However, the models demonstrated better results than a random model but also showed difficulties in predicting creditworthiness. Possible factors contributing to the model's performance are discussed in the project, along with suggestions for potential improvements. Further research is encouraged in this area to achieve better prediction accuracy. / Utvecklingen av modeller för att bedöma kunders kreditvärdighet är en väletablerad del av finanssektorn. Som en del av ett samarbete med ett svenskt kreditinstitut var målet med detta projekt att skapa en statistisk modell som kunde predicera kunders betalningsförmåga. Utöver att skapa en modell syftar projektet också till att identifiera de egenskaper hos låntagare som har hög prediktionsförmåga samt hur dessa prediktionsvariabler skiljer sig för att förutse betalningsförmågan tidigt respektive sent in i löptiden. För att undersöka detta erhölls en datamängd innehållande cirka 15 000 unika låneansökningar som godkändes mellan juli 2020 och juli 2022 från kreditinstitutet, och logistiska regressionsmodeller tillämpades med kundernas status mellan 6 och 21 månader in av löptiden som målvariabler. Modellerna visade bättre resultat än en slumpmässig modell men visade också på stora svårigheter att förutsäga kreditvärdigheten. Möjliga faktorer som bidrar till modellernas träffssäkerhet diskuteras i projektet, tillsammans med förslag på potentiella förbättringar och ytterligare forskning uppmuntras inom detta område för att uppnå bättre modeller.

This thesis describes the use of different DNA sequencing technologies to investigate the structure and function of microbial communities in two extreme environments, boreal acid sulfate soil and the terrestrial deep biosphere. The first of the two investigated environments was soils containing un-oxidized metal sulfides that are termed ‘potential acid sulfate soil’ (PASS) materials. If these materials are exposed to atmospheric oxygen by either natural phenomena (e.g., land uplift) or human activities (e.g., drainage) then the metal sulfides become oxidized and the PASS becomes acidic and is defined as an ‘acid sulfate soil’ (ASS). The resulting acid and metal release from metal sulfide oxidation can lead to severe environmental damage. Although acidophilic microorganisms capable of catalyzing acid and metal release have been identified from many sulfide mineral containing environments, the microbial community of boreal PASSs/ASSs remains unclear. This study investigated the physicochemical and microbial characteristics of PASSs and ASSs from the Risöfladan experimental field in Vasa, Finland. Sanger sequencing of 16S rRNA gene sequences of microorganisms present in the PASSs and ASSs were mostly assigned to acidophilic species and environmental clones previously identified from acid- and metal-contaminated environments. Enrichment cultures inoculated from the ASS demonstrated that the acidophilic microorganisms were responsible for catalyzing acid and metal release from PASSs/ASSs. Lastly, the study investigated how to mitigate metal sulfide oxidation and the concomitant formation of sulfuric acid by treating ASSs in situ with CaCO3 or Ca(OH)2 suspensions. The DNA sequencing still identified acidophilic microorganisms after the chemical treatments. However, the increased pH during and after treatment suggested that the activity of the acidophiles might be inhibited. This study was the first to identify the microbial community present in boreal PASSs/ASSs and suggested that treatment with basic compounds may inhibit microbial catalysis of metal sulfide dissolution. The second studied environment was the deep, dark terrestrial subsurface that is suggested to be both extremely stable and highly oligotrophic. Despite the scarcity of carbon and energy sources, the deep biosphere is estimated to constitute up to 20% of the total biomass on earth and thus, represents the largest microbial ecosystem. However, due to the difficulties of accessing this environment and our inability to cultivate the indigenous microbial populations, details of the diversity and metabolism of these communities remain largely unexplored. This study was carried out at Äspö Hard Rock Laboratory, Sweden and utilized second-generation sequencing to identify the taxonomic composition and genetic potential of planktonic and biofilm populations. Community DNA sequencing of planktonic cells from three water types at varied age and depth (‘modern marine’, ‘undefined mixed’, and ‘old saline’) showed the existence of ultra-small cells capable of passing through a 0.22 μm filter that were phylogenetically distinct communities from the >0.22 μm fraction. The reduced cell size and/or genome size suggested a potential adaptation to the oligotrophic environment in the terrestrial deep biosphere. The identified planktonic communities were dominated by Proteobacteria, Candidate divisions, unclassified archaea, and unclassified bacteria. Functional analysis of the assembled genomes showed that the planktonic population from the shallow modern marine water demonstrated a predominantly anaerobic and heterotrophic lifestyle. In contrast, the deeper, old saline water was more closely aligned with the hypothesis of a hydrogen-driven deep biosphere. Metagenomic analysis of subsurface biofilms from ‘modern marine’ and ‘old saline’ water types suggested only a subset of populations were involved in initial biofilm formation. The identified biofilm populations from both water types were distinct from the planktonic community and were suggested to be dominated by hydrogen fed, chemolithoautotrophic and diazotrophic populations.

The amount of biological sequence data is increasing rapidly, a promising development that would transform biology if we can develop methods that can analyze large-scale data efficiently and accurately. A fundamental question in evolutionary biology is building the tree of life: a reconstruction of relationships between organisms in evolutionary time. Reconstructing phylogenetic trees from molecular data is an optimization problem that involves many steps. In this dissertation, we argue that to answer long-standing phylogenetic questions with large-scale data, several challenges need to be addressed in various steps of the pipeline. One challenges is aligning large number of sequences so that evolutionarily related positions in all sequences are put in the same column. Constructing alignments is necessary for phylogenetic reconstruction, but also for many other types of evolutionary analyses. In response to this challenge, we introduce PASTA, a scalable and accurate algorithm that can align datasets with up to a million sequences. A second challenge is related to the interesting fact that various parts of the genome can have different evolutionary histories. Reconstructing a species tree from genome-scale data needs to account for these differences. A main approach for species tree reconstruction is to first reconstruct a set of ``gene trees'' from different parts of the genome, and to then summarize these gene trees into a single species tree. We argue that this approach can suffer from two challenges: reconstruction of individual gene trees from limited data can be plagued by estimation error, which translates to errors in the species tree, and also, methods that summarize gene trees are not scalable or accurate enough under some conditions. To address the first challenge, we introduce statistical binning, a method that re-estimates gene trees by grouping them into bins. We show that binning improves gene tree accuracy, and consequently the species tree accuracy. To address the second challenge, we introduce ASTRAL, a new summary method that can run on a thousand genes and a thousand species in a day and has outstanding accuracy. We show that the development of these methods has enabled biological analyses that were otherwise not possible.