Bayesian methodology for genetics of complex diseases

Genetic research of complex diseases is a challenging, but exciting, area of research. The early development of the research was limited, however, until the completion of the Human Genome and HapMap projects, along with the reduction in the cost of genotyping, which paves the way for understanding the genetic composition of complex diseases. In this thesis, we focus on the statistical methods for two aspects of genetic research: phenotype definition for diseases with complex etiology and methods for identifying potentially associated Single Nucleotide Polymorphisms (SNPs) and SNP-SNP interactions. With regard to phenotype definition for diseases with complex etiology, we firstly investigated the effects of different statistical phenotyping approaches on the subsequent analysis. In light of the findings, and the difficulties in validating the estimated phenotype, we proposed two different methods for reconciling phenotypes of different models using Bayesian model averaging as a coherent mechanism for accounting for model uncertainty. In the second part of the thesis, the focus is turned to the methods for identifying associated SNPs and SNP interactions. We review the use of Bayesian logistic regression with variable selection for SNP identification and extended the model for detecting the interaction effects for population based case-control studies. In this part of study, we also develop a machine learning algorithm to cope with the large scale data analysis, namely modified Logic Regression with Genetic Program (MLR-GEP), which is then compared with the Bayesian model, Random Forests and other variants of logic regression.

Bayesian survival analysis using gene expression

This thesis developed and applied Bayesian models for the analysis of survival data. The gene expression was considered as explanatory variables within the Bayesian survival model which can be considered the new contribution in the analysis of such data. The censoring factor that is inherent of survival data has also been addressed in terms of its impact on the fitting of a finite mixture of Weibull distribution with and without covariates. To investigate this, simulation study were carried out under several censoring percentages. Censoring percentage as high as 80% is acceptable here as the work involved high dimensional data. Lastly the Bayesian model averaging approach was developed to incorporate model uncertainty in the prediction of survival.

Identifying the spatial scale of land use that most strongly influences overall river ecosystem health score

Catchment and riparian degradation has resulted in declining ecosystem health of streams worldwide. With restoration a priority in many regions, there is an increasing interest in the scale at which land use influences stream ecosystem health. Our goal was to use a substantial data set collected as part of a monitoring program (the Southeast Queensland, Australia, Ecological Health Monitoring Program data set, collected at 116 sites over six years) to identify the spatial scale of land use, or the combination of spatial scales, that most strongly influences overall ecosystem health. In addition, we aimed to determine whether the most influential scale differed for different aspects of ecosystem health. We used linear-mixed models and a Bayesian model-averaging approach to generate models for the overall aggregated ecosystem health score and for each of the five component indicators (fish, macroinvertebrates, water quality, nutrients, and ecosystem processes) that make up the score. Dense forest close to the survey site, mid-dense forest in the hydrologically active nearstream areas of the catchment, urbanization in the riparian buffer, and tree cover at the reach scale were all significant in explaining ecosystem health, suggesting an overriding influence of forest cover, particularly close to the stream. Season and antecedent rainfall were also important explanatory variables, with some land-use variables showing significant seasonal interactions. There were also differential influences of land use for each of the component indicators. Our approach is useful given that restoring general ecosystem health is the focus of many stream restoration projects; it allowed us to predict the scale and catchment position of restoration that would result in the greatest improvement of ecosystem health in the regions streams and rivers. The models we generated suggested that good ecosystem health can be maintained in catchments where 80% of hydrologically active areas in close proximity to the stream have mid-dense forest cover and moderate health can be obtained with 60% cover.

Eliciting and encoding expert knowledge on variable selection into classical or Bayesian species distribution models

The quality of species distribution models (SDMs) relies to a large degree on the quality of the input data, from bioclimatic indices to environmental and habitat descriptors (Austin, 2002). Recent reviews of SDM techniques, have sought to optimize predictive performance e.g. Elith et al., 2006. In general SDMs employ one of three approaches to variable selection. The simplest approach relies on the expert to select the variables, as in environmental niche models Nix, 1986 or a generalized linear model without variable selection (Miller and Franklin, 2002). A second approach explicitly incorporates variable selection into model fitting, which allows examination of particular combinations of variables. Examples include generalized linear or additive models with variable selection (Hastie et al. 2002); or classification trees with complexity or model based pruning (Breiman et al., 1984, Zeileis, 2008). A third approach uses model averaging, to summarize the overall contribution of a variable, without considering particular combinations. Examples include neural networks, boosted or bagged regression trees and Maximum Entropy as compared in Elith et al. 2006. Typically, users of SDMs will either consider a small number of variable sets, via the first approach, or else supply all of the candidate variables (often numbering more than a hundred) to the second or third approaches. Bayesian SDMs exist, with several methods for eliciting and encoding priors on model parameters (see review in Low Choy et al. 2010). However few methods have been published for informative variable selection; one example is Bayesian trees (O’Leary 2008). Here we report an elicitation protocol that helps makes explicit a priori expert judgements on the quality of candidate variables. This protocol can be flexibly applied to any of the three approaches to variable selection, described above, Bayesian or otherwise. We demonstrate how this information can be obtained then used to guide variable selection in classical or machine learning SDMs, or to define priors within Bayesian SDMs.

Bagging and the Bayesian Bootstrap

Bagging is a method of obtaining more ro- bust predictions when the model class under consideration is unstable with respect to the data, i.e., small changes in the data can cause the predicted values to change significantly. In this paper, we introduce a Bayesian ver- sion of bagging based on the Bayesian boot- strap. The Bayesian bootstrap resolves a the- oretical problem with ordinary bagging and often results in more efficient estimators. We show how model averaging can be combined within the Bayesian bootstrap and illustrate the procedure with several examples.

Improving parameter learning of Bayesian nets from incomplete data

This paper addresses the estimation of parameters of a Bayesian network from incomplete data. The task is usually tackled by running the Expectation-Maximization (EM) algorithm several times in order to obtain a high log-likelihood estimate. We argue that choosing the maximum log-likelihood estimate (as well as the maximum penalized log-likelihood and the maximum a posteriori estimate) has severe drawbacks, being affected both by overfitting and model uncertainty. Two ideas are discussed to overcome these issues: a maximum entropy approach and a Bayesian model averaging approach. Both ideas can be easily applied on top of EM, while the entropy idea can be also implemented in a more sophisticated way, through a dedicated non-linear solver. A vast set of experiments shows that these ideas produce significantly better estimates and inferences than the traditional and widely used maximum (penalized) log-likelihood and maximum a posteriori estimates. In particular, if EM is adopted as optimization engine, the model averaging approach is the best performing one; its performance is matched by the entropy approach when implemented using the non-linear solver. The results suggest that the applicability of these ideas is immediate (they are easy to implement and to integrate in currently available inference engines) and that they constitute a better way to learn Bayesian network parameters.

Validation des modèles statistiques tenant compte des variables dépendantes du temps en prévention primaire des maladies cérébrovasculaires

L’intérêt principal de cette recherche porte sur la validation d’une méthode statistique en pharmaco-épidémiologie. Plus précisément, nous allons comparer les résultats d’une étude précédente réalisée avec un devis cas-témoins niché dans la cohorte utilisé pour tenir compte de l’exposition moyenne au traitement : – aux résultats obtenus dans un devis cohorte, en utilisant la variable exposition variant dans le temps, sans faire d’ajustement pour le temps passé depuis l’exposition ; – aux résultats obtenus en utilisant l’exposition cumulative pondérée par le passé récent ; – aux résultats obtenus selon la méthode bayésienne. Les covariables seront estimées par l’approche classique ainsi qu’en utilisant l’approche non paramétrique bayésienne. Pour la deuxième le moyennage bayésien des modèles sera utilisé pour modéliser l’incertitude face au choix des modèles. La technique utilisée dans l’approche bayésienne a été proposée en 1997 mais selon notre connaissance elle n’a pas été utilisée avec une variable dépendante du temps. Afin de modéliser l’effet cumulatif de l’exposition variant dans le temps, dans l’approche classique la fonction assignant les poids selon le passé récent sera estimée en utilisant des splines de régression. Afin de pouvoir comparer les résultats avec une étude précédemment réalisée, une cohorte de personnes ayant un diagnostique d’hypertension sera construite en utilisant les bases des données de la RAMQ et de Med-Echo. Le modèle de Cox incluant deux variables qui varient dans le temps sera utilisé. Les variables qui varient dans le temps considérées dans ce mémoire sont iv la variable dépendante (premier évènement cérébrovasculaire) et une des variables indépendantes, notamment l’exposition

Identification of Putative Wintering Areas and Ecological Determinants of Population Dynamics of Common House-Martin (Delichon urbicum) and Common Swift (Apus apus) Breeding in Northern Italy

To identify the causes of population decline in migratory birds, researchers must determine the relative influence of environmental changes on population dynamics while the birds are on breeding grounds, wintering grounds, and en route between the two. This is problematic when the wintering areas of specific populations are unknown. Here, we first identified the putative wintering areas of Common House-Martin (Delichon urbicum) and Common Swift (Apus apus) populations breeding in northern Italy as those areas, within the wintering ranges of these species, where the winter Normalized Difference Vegetation Index (NDVI), which may affect winter survival, best predicted annual variation in population indices observed in the breeding grounds in 1992–2009. In these analyses, we controlled for the potentially confounding effects of rainfall in the breeding grounds during the previous year, which may affect reproductive success; the North Atlantic Oscillation Index (NAO), which may account for climatic conditions faced by birds during migration; and the linear and squared term of year, which account for nonlinear population trends. The areas thus identified ranged from Guinea to Nigeria for the Common House-Martin, and were located in southern Ghana for the Common Swift. We then regressed annual population indices on mean NDVI values in the putative wintering areas and on the other variables, and used Bayesian model averaging (BMA) and hierarchical partitioning (HP) of variance to assess their relative contribution to population dynamics. We re-ran all the analyses using NDVI values at different spatial scales, and consistently found that our population of Common House-Martin was primarily affected by spring rainfall (43%–47.7% explained variance) and NDVI (24%–26.9%), while the Common Swift population was primarily affected by the NDVI (22.7%–34.8%). Although these results must be further validated, currently they are the only hypotheses about the wintering grounds of the Italian populations of these species, as no Common House-Martin and Common Swift ringed in Italy have been recovered in their wintering ranges.

Modeling Habitat Associations for the Common Loon (Gavia immer) at Multiple Scales in Northeastern North America

Common Loon (Gavia immer) is considered an emblematic and ecologically important example of aquatic-dependent wildlife in North America. The northern breeding range of Common Loon has contracted over the last century as a result of habitat degradation from human disturbance and lakeshore development. We focused on the state of New Hampshire, USA, where a long-term monitoring program conducted by the Loon Preservation Committee has been collecting biological data on Common Loon since 1976. The Common Loon population in New Hampshire is distributed throughout the state across a wide range of lake-specific habitats, water quality conditions, and levels of human disturbance. We used a multiscale approach to evaluate the association of Common Loon and breeding habitat within three natural physiographic ecoregions of New Hampshire. These multiple scales reflect Common Loon-specific extents such as territories, home ranges, and lake-landscape influences. We developed ecoregional multiscale models and compared them to single-scale models to evaluate model performance in distinguishing Common Loon breeding habitat. Based on information-theoretic criteria, there is empirical support for both multiscale and single-scale models across all three ecoregions, warranting a model-averaging approach. Our results suggest that the Common Loon responds to both ecological and anthropogenic factors at multiple scales when selecting breeding sites. These multiscale models can be used to identify and prioritize the conservation of preferred nesting habitat for Common Loon populations.

How large is congressional dependence in agriculture. Bayesian inference about ‘scale’ and ‘scope’ in measuring a spatial externality

The political economy literature on agriculture emphasizes influence over political outcomes via lobbying conduits in general, political action committee contributions in particular and the pervasive view that political preferences with respect to agricultural issues are inherently geographic. In this context, ‘interdependence’ in Congressional vote behaviour manifests itself in two dimensions. One dimension is the intensity by which neighboring vote propensities influence one another and the second is the geographic extent of voter influence. We estimate these facets of dependence using data on a Congressional vote on the 2001 Farm Bill using routine Markov chain Monte Carlo procedures and Bayesian model averaging, in particular. In so doing, we develop a novel procedure to examine both the reliability and the consequences of different model representations for measuring both the ‘scale’ and the ‘scope’ of spatial (geographic) co-relations in voting behaviour.

Parametric preference functionals under risk in the gain domain: a Bayesian analysis

The performance of rank dependent preference functionals under risk is comprehensively evaluated using Bayesian model averaging. Model comparisons are made at three levels of heterogeneity plus three ways of linking deterministic and stochastic models: the differences in utilities, the differences in certainty equivalents and contextualutility. Overall, the"bestmodel", which is conditional on the form of heterogeneity is a form of Rank Dependent Utility or Prospect Theory that cap tures the majority of behaviour at both the representative agent and individual level. However, the curvature of the probability weighting function for many individuals is S-shaped, or ostensibly concave or convex rather than the inverse S-shape commonly employed. Also contextual utility is broadly supported across all levels of heterogeneity. Finally, the Priority Heuristic model, previously examined within a deterministic setting, is estimated within a stochastic framework, and allowing for endogenous thresholds does improve model performance although it does not compete well with the other specications considered.

Untangling associations between chironomid taxa in Neotropical streams using local and landscape filters

1. Analyses of species association have major implications for selecting indicators for freshwater biomonitoring and conservation, because they allow for the elimination of redundant information and focus on taxa that can be easily handled and identified. These analyses are particularly relevant in the debate about using speciose groups (such as the Chironomidae) as indicators in the tropics, because they require difficult and time-consuming analysis, and their responses to environmental gradients, including anthropogenic stressors, are poorly known. 2. Our objective was to show whether chironomid assemblages in Neotropical streams include clear associations of taxa and, if so, how well these associations could be explained by a set of models containing information from different spatial scales. For this, we formulated a priori models that allowed for the influence of local, landscape and spatial factors on chironomid taxon associations (CTA). These models represented biological hypotheses capable of explaining associations between chironomid taxa. For instance, CTA could be best explained by local variables (e.g. pH, conductivity and water temperature) or by processes acting at wider landscape scales (e.g. percentage of forest cover). 3. Biological data were taken from 61 streams in Southeastern Brazil, 47 of which were in well-preserved regions, and 14 of which drained areas severely affected by anthropogenic activities. We adopted a model selection procedure using Akaike`s information criterion to determine the most parsimonious models for explaining CTA. 4. Applying Kendall`s coefficient of concordance, seven genera (Tanytarsus/Caladomyia, Ablabesmyia, Parametriocnemus, Pentaneura, Nanocladius, Polypedilum and Rheotanytarsus) were identified as associated taxa. The best-supported model explained 42.6% of the total variance in the abundance of associated taxa. This model combined local and landscape environmental filters and spatial variables (which were derived from eigenfunction analysis). However, the model with local filters and spatial variables also had a good chance of being selected as the best model. 5. Standardised partial regression coefficients of local and landscape filters, including spatial variables, derived from model averaging allowed an estimation of which variables were best correlated with the abundance of associated taxa. In general, the abundance of the associated genera tended to be lower in streams characterised by a high percentage of forest cover (landscape scale), lower proportion of muddy substrata and high values of pH and conductivity (local scale). 6. Overall, our main result adds to the increasing number of studies that have indicated the importance of local and landscape variables, as well as the spatial relationships among sampling sites, for explaining aquatic insect community patterns in streams. Furthermore, our findings open new possibilities for the elimination of redundant data in the assessment of anthropogenic impacts on tropical streams.

Neural network ensemble : evaluation of aggregation algorithms for electricity demand forecasting

This paper examines and analyzes different aggregation algorithms to improve accuracy of forecasts obtained using neural network (NN) ensembles. These algorithms include equal-weights combination of Best NN models, combination of trimmed forecasts, and Bayesian Model Averaging (BMA). The predictive performance of these algorithms are evaluated using Australian electricity demand data. The output of the aggregation algorithms of NN ensembles are compared with a Naive approach. Mean absolute percentage error is applied as the performance index for assessing the quality of aggregated forecasts. Through comprehensive simulations, it is found that the aggregation algorithms can significantly improve the forecasting accuracies. The BMA algorithm also demonstrates the best performance amongst aggregation algorithms investigated in this study.

Relationship between long-term environmental fluctuations and diving effort of female Australian fur seals

For predators foraging within spatially and temporally heterogeneous marine ecosystems, environmental fluctuations can alter prey availability. Using the proportion of time spent diving and foraging trip duration as proxies of foraging effort, a multi-year dataset was used to assess the response of 58 female Australian fur seals Arctocephalus pusillus doriferus to interannual environmental fluctuations. Multiple environmental indices (remotely sensed ocean colour data and numerical weather predictions) were assessed for their influence on inter-annual variations in the proportion of time spent diving and trip duration. Model averaging revealed strong evidence for relationships between 4 indices and the proportion of time spent diving. There was a positive relationship with effort and 2 yr-lagged spring sea-surface temperature, current winter zonal wind and southern oscillation index, while a negative relationship was found with 2 yr-lagged spring zonal wind. Additionally, a positive relationship was found between foraging trip duration and 1 yr-lagged spring surface chlorophyll a. These results suggest that environmental fluctuations may influence prey availability by affecting the survival and recruitment of prey at the larval and post-larval phases while also affecting current distribution of adult prey.

Examining performance of aggregation algorithms for neural network-based electricity demand forecasting

The aim of this research is to examine the efficiency of different aggregation algorithms to the forecasts obtained from individual neural network (NN) models in an ensemble. In this study an ensemble of 100 NN models are constructed with a heterogeneous architecture. The outputs from NN models are combined by three different aggregation algorithms. These aggregation algorithms comprise of a simple average, trimmed mean, and a Bayesian model averaging. These methods are utilized with certain modifications and are employed on the forecasts obtained from all individual NN models. The output of the aggregation algorithms is analyzed and compared with the individual NN models used in NN ensemble and with a Naive approach. Thirty-minutes interval electricity demand data from Australian Energy Market Operator (AEMO) and the New York Independent System Operator's web site (NYISO) are used in the empirical analysis. It is observed that the aggregation algorithm perform better than many of the individual NN models. In comparison with the Naive approach, the aggregation algorithms exhibit somewhat better forecasting performance.