Validation-Based Sparse Gaussian Process Classifier Design

Gaussian processes (GPs) are promising Bayesian methods for classification and regression problems. Design of a GP classifier and making predictions using it is, however, computationally demanding, especially when the training set size is large. Sparse GP classifiers are known to overcome this limitation. In this letter, we propose and study a validation-based method for sparse GP classifier design. The proposed method uses a negative log predictive (NLP) loss measure, which is easy to compute for GP models. We use this measure for both basis vector selection and hyperparameter adaptation. The experimental results on several real-world benchmark data sets show better orcomparable generalization performance over existing methods.

Modeling ecological and evolutionary processes in spatially structured populations

Many species inhabit fragmented landscapes, resulting either from anthropogenic or from natural processes. The ecological and evolutionary dynamics of spatially structured populations are affected by a complex interplay between endogenous and exogenous factors. The metapopulation approach, simplifying the landscape to a discrete set of patches of breeding habitat surrounded by unsuitable matrix, has become a widely applied paradigm for the study of species inhabiting highly fragmented landscapes. In this thesis, I focus on the construction of biologically realistic models and their parameterization with empirical data, with the general objective of understanding how the interactions between individuals and their spatially structured environment affect ecological and evolutionary processes in fragmented landscapes. I study two hierarchically structured model systems, which are the Glanville fritillary butterfly in the Åland Islands, and a system of two interacting aphid species in the Tvärminne archipelago, both being located in South-Western Finland. The interesting and challenging feature of both study systems is that the population dynamics occur over multiple spatial scales that are linked by various processes. My main emphasis is in the development of mathematical and statistical methodologies. For the Glanville fritillary case study, I first build a Bayesian framework for the estimation of death rates and capture probabilities from mark-recapture data, with the novelty of accounting for variation among individuals in capture probabilities and survival. I then characterize the dispersal phase of the butterflies by deriving a mathematical approximation of a diffusion-based movement model applied to a network of patches. I use the movement model as a building block to construct an individual-based evolutionary model for the Glanville fritillary butterfly metapopulation. I parameterize the evolutionary model using a pattern-oriented approach, and use it to study how the landscape structure affects the evolution of dispersal. For the aphid case study, I develop a Bayesian model of hierarchical multi-scale metapopulation dynamics, where the observed extinction and colonization rates are decomposed into intrinsic rates operating specifically at each spatial scale. In summary, I show how analytical approaches, hierarchical Bayesian methods and individual-based simulations can be used individually or in combination to tackle complex problems from many different viewpoints. In particular, hierarchical Bayesian methods provide a useful tool for decomposing ecological complexity into more tractable components.

Semi-Supervised Classification Using Sparse Gaussian Process Regression

Gaussian Processes (GPs) are promising Bayesian methods for classification and regression problems. They have also been used for semi-supervised learning tasks. In this paper, we propose a new algorithm for solving semi-supervised binary classification problem using sparse GP regression (GPR) models. It is closely related to semi-supervised learning based on support vector regression (SVR) and maximum margin clustering. The proposed algorithm is simple and easy to implement. It gives a sparse solution directly unlike the SVR based algorithm. Also, the hyperparameters are estimated easily without resorting to expensive cross-validation technique. Use of sparse GPR model helps in making the proposed algorithm scalable. Preliminary results on synthetic and real-world data sets demonstrate the efficacy of the new algorithm.

Diversidade genética em um grupo de anuros em miniatura endêmico da Mata Atlântica brasileira (Euparkerella Griffiths 1959).

A Mata Atlântica (MA) está entre as regiões com maior biodiversidade e mais ameaçadas do planeta. Esforços em diversas áreas do conhecimento têm sido feitos para que se tenha uma estimativa mais refinada da diversidade existente e sua organização ao longo do bioma. O crescente número de estudos que buscam reconstituir a história da diversificação da MA apontam para um cenário espacial e temporal complexo, havendo ainda uma lacuna no conhecimento dos processos em pequena escala. Vertebrados em miniatura têm se mostrado uma boa ferramenta para estudos de processos evolutivos em pequena escala. Assim, o gênero Euparkerella, endêmico de uma pequena região da MA dos Estados do Rio de Janeiro (RJ) e Espírito Santo (ES), foi escolhido como modelo para este estudo. No primeiro capítulo buscou-se descrever a diversidade existente dentro do gênero a partir de uma filogenia molecular. Para isso, utilizaram-se métodos bayesianos para gerar genealogias de genes e de espécies a partir de um fragmento de gene mitocondrial e quatro fragmentos de genes nucleares. Os resultados obtidos apontaram para uma grande diversidade críptica no gênero. Foram identificadas seis unidades evolutivas significativamente divergentes para o RJ: duas em Euparkerella cochranae, três em Euparkerella brasiliensis, e Euparkerella sp.. A espécie mais basal recuperada foi Euparkerella robusta, do ES, e estimou-se o início da diversificação do gênero para o final do Mioceno. O segundo capítulo descreve onze marcadores de microssatélites desenvolvidos para Euparkerella brasiliensis através do método de pirosequenciamento de nova geração 454. No terceiro capítulo estudou-se apenas uma unidade evolutiva, Euparkerella brasiliensis da área dos Três Picos/ RJ. A partir de marcadores de evolução rápida (microssatélites) e lenta (sequências de DNA) buscou-se compreender a estrutura e a dinâmica populacional desta unidade evolutiva em uma área bastante pequena (aprox. 20 km) sob influência de um gradiente ambiental altitudinal (40 m 1000 m). Foram identificadas, a partir dos microssatélites, duas subpopulações geneticamente distintas nas bordas do gradiente. O fluxo gênico se deu predominantemente das bordas para a zona de contato, onde foi observado o maior efetivo populacional. Tais resultados indicam que pequenas variações ambientais podem atuar no isolamento populacional em Euparkerella e corroboram o padrão de formas microendêmicas identificadas na filogenia. Futuros estudos devem ser feitos no sentido de buscar caracterizar morfologicamente as unidades evolutivas aqui identificadas; preencher as lacunas amostrais, especialmente no ES; e descrever os processos que atuam em pequena escala nas zonas de contato entre as unidades evolutivas e fatores limitantes a distribuição das mesmas.

Phylogenetic relationships of Drosophila melanogaster species group deduced from spacer regions of histone gene H2A-H2B

Nucleotide sequences of the spacer region of the histone gene H2A-H2B from 36 species of Drosophila melanogaster species group were determined. The phylogenetic trees were reconstructed with maximum parsimony, maximum likelihood, and Bayesian methods by u

Phylogenetic relationships of the macaques (Cercopithecidae : Macaca), inferred from mitochondrial DNA sequences

To study the phylogenetic relationships of the macaques, five gene fragments were sequenced from 40 individuals of eight species: Macaca mulatta, M. cyclopis, M. fascicularis, M. arctoides, M. assamensis, M. thibetana, M. silenus, and M. leonina. In addition, sequences of M. sylvanus were obtained from Genbank. A baboon was used as the outgroup. The phylogenetic trees were constructed using maximum-parsimony and Bayesian methods. Because five gene fragments were from the mitochondrial genome and were inherited as a single entity without recombination, we combined the five genes into a single analysis. The parsimony bootstrap proportions we obtained were higher than those from earlier studies based on the combined mtDNA dataset. Excluding M. arctoides, our results are generally consistent with the classification of Delson (1980). Our phylogenetic analyses agree with earlier studies suggesting that the mitochondrial lineages of M. arctoides share a close evolutionary relationship with the mitochondrial lineages of the fascicularis group of macaques (and M. fascicularis, specifically). M. mulatta (with respect to M. cyclopis), M. assamensis assamensis (with respect to M. thibetana), and M. leonina (with respect to M. silenus) are paraphyletic based on our analysis of mitochondrial genes.

Phylogeny and biogeography of the Petaurista philippensis complex (Rodentia : Sciuridae), inter- and intraspecific relationships inferred from molecular and morphometric analysis

With modified DNA extraction and Purification protocols, the complete cytochrome b gene sequences (1140 bp) were determined from degraded museum specimens. Molecular analysis and morphological examination of cranial characteristics of the giant flying squirrels of Petaurista philippensis complex (P. grandis, P. hainana, and P. yunanensis) and other Petaurista species yielded new insights into long-standing controversies in the Petaurista systematics. Patterns of genetic variations and morphological differences observed in this study indicate that P. hainana, P. albiventer, and P. yunanensis can be recognized as distinct species, and P. grandis and P. petaurista are conspecific populations. Phylogenetic relationships reconstructed by using parsimony, likelihood, and Bayesian methods reveal that, with P. leucogenys as the basal branch, all Petaurista groups formed two distinct clades. Petaurista philippensis, P. hainana, P. yunanensis, and P. albiventer are clustered in the same clade, while P. grandis shows a close relationship to P. petaurista. Deduced divergence times based on Bayesian analysis and the transversional substitution at the third codon suggest that the retreating of glaciers and upheavals or movements of tectonic plates in the Pliocene-Pleistocene were the major factors responsible for the present geographical distributions of Petaurista groups. (c) 2005 Elsevier Inc. All rights reserved.

Phylogeny and speciation of the eastern Asian cyprinid genus Sarcocheilichthys

The genus Sarcocheilichthys is a group of small cyprinid fishes comprising 10 species/sub-species widely distributed in East Asia, which represents a valuable model for understanding the speciation of freshwater fishes in East Asia. In the present study, the molecular phylogenetic relationship of the genus Sarcocheilichthys was investigated using a 1140 bp section of the mitochondrial cytochrome b gene. Two different tree-building methods, maximum parsimony (MP) and Bayesian methods, yielded trees with almost the same topology, yielding high bootstrap values or posterior probabilities. The results showed that the genus Sarcocheilichthys consists of two large clades, clades I and II. Clade I contains Sarcocheilichthys lacustris, Sarcocheilichthys sinensis and Sarcocheilichthys parvus, with S. parvus at a basal position. In clade II, Sarcocheilichthys variegatus microoculus is at a basal position; samples of the widespread species, Sarcocheilichthys nigripinnis, form a large subclade containing another valid species Sarcocheilichthys czerskii. Sarcocheilichthys kiangsiensis is retained at an intermediate position. Since S. czerskii is a valid species in the S. nigripinnis clade, remaining samples of S. nigripinnis form a paraphyly. This speciation process is attributed to geographical isolation and special environmental conditions experienced by S. czerskii and stable environments experienced by the other S. nigripinnis populations. This type of speciation process was suggested to be very common. Samples of Sarcocheilichthys sinensis sinensis and Sarcocheilichthys sinensis fukiensis that did not form their own monophyletic groups suggest an early stage of speciation and support their sub-species status. Molecular clock analysis indicates that the two major lineages of the genus Sarcocheilichthys, clades I and II diverged c. 8.89 million years ago (mya). Sarcocheilichthys v. microoculus from Japan probably diverged 4.78 mya from the Chinese group. The northern-southern clades of S. nigripinnis began to diverge c. 2.12 mya, while one lineage of S. nigripinnis evolved into a new species, S. czerski, c. 0.34 mya. (C) 2008 The Authors Journal compilation (C) 2008 The Fisheries Society of the British Isles.

Investigations into the perplexing interrelationship of the Genus Takifugu Abe, 1949 (Tetraodontiformes, Tetraodontidae)

The phylogenetic relationships within the genus Takifugu Abe, 1949 (Tetraodontiformes, Tetraodontidae) remain unresolved. Because of the use of Takifugu as model organisms, the resolution of these relationships is crucial for the interpretation of evolutionary trends in biology. Pufferfishes of this genus are comprised of a comparatively small number of species and are mainly distributed along the coastal region of the western part of the Sea of Japan and the coastline of China. Mitochondrial gene sequences were employed to test the phylogenetic hypotheses within the genus. Seventeen species of the genus were examined. Molecular phylogenetic trees were constructed using the maximum parsimony, neighbor-joining, maximum likelihood and Bayesian methods. Our hypothesis of internal relationships within the genus differs from previous hypotheses. Our results indicate that (1) the genus Takifugu is a monophyletic assemblage; (2) the genus is divided into 6 subgroups based on the molecular data; and (3) there is low genetic diversity among the species within this genus. In addition, speciation within Takifugu appears to be driven by hybridization and isolation by distribution. Our results also suggested that the taxonomy in the genus should be clarified based on both molecular and morphological data.

Phylogenetic relationships of Cyprinidae (Teleostei : Cypriniformes) inferred from the partial S6K1 gene sequences and implication of indel sites in intron 1

The family Cyprinidae is widely distributed in East Asia, and has the important phylogenetic significance in the fish evolution. In this study, the 5' end partial sequences (containing exon 1, exon 2 and indel 1) of S6K1 gene were obtained from 30 representative species in Cyprinidae and outgroup using PCR amplification and sequencing. The phylogenetic relationships of Cyprinidae were reconstructed with neighbor joining (NJ), maximum parsimony (MP), maximum likelihood (ML), and Bayesian methods. Myxocyprinus asiaticus (Catostomidae) was assigned to the outgroup taxon. Similar phylogenetic relationships within the family Cyprinidae were achieved with the four analyses. Leuciscini and Barbini were monophyletic lineages respectively with the high nodal supports. Leuciscini comprises Hypophthalmichthyinae, Xenocyprinae, Cultrinae, Gobioninae, Acheilognathinae and East Asian species of Leuciscinae and Danioninae. Monophyly of East Asian clade was supported with high nodal support. Barbini comprises Schizothoracinae, Barbinae, Cyprininae and Labeoninae. The monophyletic lineage consisting of Danio rerio, D. myersi, and Rasbora trilineata was basal in the tree. In addition, the large fragment indels in intron 1 were analyzed to improve the understanding of Cyprinidae relationships. The results showed that the large fragment indels were correlated with the relations among species. Some conserved regions in intron 1 were thought to be involved in the functional regulation. However, no correlation was found between sequence variations and species characteristic size.

Phylogeny of diplozoids in five genera of the subfamily diplozoinae palombi, 1949 as inferred from ITS-2 rDNA sequences

The phylogenetic relationship of 5 genera, i.e. Diplozoon Nordmann, 1832, Paradiplozoon Achmerov, 1974, Inustiatus Khotenovsky, 1978, Sindiplozoon Khotenovsky, 1981, and Eudiplozoon Khotenovsky, 1985 in the subfamily Diplozoinae Palombi, 1949 (Monogenea, Polyopisthocotylea) was inferred from rDNA ITS-2 region using neighbour-joining (NJ), maximum likelihood (ML) and Bayesian methods. The phylogenetic trees produced by using NJ, ML and Bayesian methods exhibit essentially the same topology. Surprisingly, freshwater species of Paradiplozoon from Europe clustered together with species of Diplozoon, but separated from Chinese Paradiplozoon species. The results of molecular phylogeny and lower level of divergence (4(.)1-15(.)7%) in ITS-2 rDNA among Paradiplozoon from Europe and Diplozoon and, on the other hand, high level of divergence (45(.)3-53(.)7%) among Paradiplozoon species from Europe and China might indicate the non-monophyletic origin of the genus Paradiplozoon. Also, the generic status of European Paradiplozoon needs to be revised. The species of Paradiplozoon in China is a basal group in Diplozoinae as revealed by NJ and Bayesian methods, and Sindiplozoon appears to be closely related to European Paradiplozoon and Diplozoon. with their relationship to Eudiplozoon and Inustiatus being unresolved.

Biogeography and molecular phylogeny of the genus Schizothorax (Teleostei : Cyprinidae) in China inferred from cytochrome b sequences

Aim: To test a vicariant speciation hypothesis derived from geological evidence of large-scale changes in drainage patterns in the late Miocene that affected the drainages in the south-eastern Tibetan Plateau. Location: The Tibetan Plateau and adjacent areas. Methods: The cytochrome b DNA sequences of 30 species of the genus Schizothorax from nine different river systems were analysed. These DNA sequences were analysed using parsimony, maximum likelihood and Bayesian methods. The approximately unbiased and Shimodaira-Hasegawa tests were applied to evaluate the statistical significance of the shortest trees relative to alternative hypotheses. Dates of divergences between lineages were estimated using the nonparametric rate smoothing method, and confidence intervals of dates were obtained by parametric bootstrapping. Results: The phylogenetic relationships recovered from molecular data were inconsistent with traditional taxonomy, but apparently reflected geographical associations with rivers. Within the genus Schizothorax, we observed a divergence between the lineages from the Irrawaddy-Lhuit and Tsangpo-Parlung rivers, and tentatively dated this vicariant event back to the late Miocene (7.3-6.8 Ma). We also observed approximately simultaneous geographical splits within drainages of the south-eastern Tibetan Plateau, the Irrawaddy, the Yangtze and the Mekong-Salween rivers in the late Miocene (7.1-6.2 Ma). Main conclusions: Our molecular evidence tentatively highlights the importance of palaeoriver connections and the uplift of the Tibetan Plateau in understanding the evolution of the genus Schizothorax. Molecular estimates of divergence times allowed us to date these vicariant scenarios back to the late Miocene, which agrees with geological suggestions for the separation of these drainages caused by tectonic uplift in south-eastern Tibet. Our results indicated the substantial role of vicariant-based speciation in shaping the current distribution pattern of the genus Schizothorax.

Comparison of evolutionary rates in the mitochondrial DNA cytochrome b gene and control region and their implications for phylogeny of the Cobitoidea (Teleostei : Cypriniformes)

It is widely accepted that mitochondrial DNA (mtDNA) control region evolves faster than protein encoding genes with few exceptions. In the present study, we sequenced the mitochondrial cytochrome b gene (cyt b) and control region (CR) and compared their rates in 93 specimens representing 67 species of loaches and some related taxa in the Cobitoidea (Order Cypriniformes). The results showed that sequence divergences of the CR were broadly higher than those of the cyt b (about 1.83 times). However, in considering only closely related species, CR sequence evolution was slower than that of cyt b gene (ratio of CR/cyt b is 0.78), a pattern that is found to be very common in Cypriniformes. Combined data of the cyt b and CR were used to estimate the phylogenetic relationship of the Cobitoidea by maximum parsimony, neighbor-joining, and Bayesian methods. With Cyprinus carpio and Danio rerio as outgroups, three analyses identified the same four lineages representing four subfamilies of loaches, with Botiinae on the basal-most clade. The phylogenctic relationship of the Cobitoidea was ((Catostomidae + Gyrinocheilidae) + (Botiinae + (Balitorinae + (Cobitinae + Nemacheilinae)))), which indicated that Sawada's Cobitidae (including Cobitinae and Botiinae) was not monophyletic. Our molecular phylogenetic analyses are in very close agreement with the phylogenetic results based on the morphological data proposed by Nalbant and Bianco, wherein these four subfamilies were elevated to the family level as Botiidae, Balitoridae, Cobitidae, and Nemacheilidae. (c) 2005 Elsevier Inc. All rights reserved.

Phylogeny of the East Asian botiine loaches (Cypriniformes, Botiidae) inferred from mitochondrial cytochrome b gene sequences

The Botiinae have traditionally represented a subfamily of the Cobitidae. At present, the classification and phylogenetic relationships of the Botiinae are controversial. To address systematic and phylogenetic questions concerning this group, we sequenced the complete cytochrome b gene from 34 samples, of which 24 represented 13 species of the East Asian botiine fishes, while the other 10 were non-botiine loach species. For the 1140 bp sequences determined, 494 sites were variable ones, of which 424 were parsimony informative. With Myxocyprinus asiaticus as an outgroup, molecular phylogenetic trees were constructed using the neighbor-joining, maximum parsimony, maximum likelihood and Bayesian methods. All molecular phylogenetic trees revealed that botiine fishes form a monophyletic group and are distantly related to other loaches, suggesting that the Botiinae should be placed in their own family. Within the Botiinae, there are three genera; Botia, Parabotia, and Leptobotia, each genus forming a monophyletic group, with the genus Botia as the most ancestral split. Our molecular results are in agreement with morphological analyses of botiines, suggesting that Botia is the ancestral genus, while Leptobotia and Parabotia were resolved as more derived sister groups.