Approximate Bayesian Computation for astronomical model analysis : a case study in galaxy demographics and morphological transformation at high redshift

Approximate Bayesian Computation’ (ABC) represents a powerful methodology for the analysis of complex stochastic systems for which the likelihood of the observed data under an arbitrary set of input parameters may be entirely intractable – the latter condition rendering useless the standard machinery of tractable likelihood-based, Bayesian statistical inference [e.g. conventional Markov chain Monte Carlo (MCMC) simulation]. In this paper, we demonstrate the potential of ABC for astronomical model analysis by application to a case study in the morphological transformation of high-redshift galaxies. To this end, we develop, first, a stochastic model for the competing processes of merging and secular evolution in the early Universe, and secondly, through an ABC-based comparison against the observed demographics of massive (Mgal > 1011 M⊙) galaxies (at 1.5 < z < 3) in the Cosmic Assembly Near-IR Deep Extragalatic Legacy Survey (CANDELS)/Extended Groth Strip (EGS) data set we derive posterior probability densities for the key parameters of this model. The ‘Sequential Monte Carlo’ implementation of ABC exhibited herein, featuring both a self-generating target sequence and self-refining MCMC kernel, is amongst the most efficient of contemporary approaches to this important statistical algorithm. We highlight as well through our chosen case study the value of careful summary statistic selection, and demonstrate two modern strategies for assessment and optimization in this regard. Ultimately, our ABC analysis of the high-redshift morphological mix returns tight constraints on the evolving merger rate in the early Universe and favours major merging (with disc survival or rapid reformation) over secular evolution as the mechanism most responsible for building up the first generation of bulges in early-type discs.

Class notes : 10 principles of good BPM

The question concerning what makes for good BPM is often raised. A recent call from Paul Harmon on the BPTrends Discussion LinkedIN Group for key issues in BPM received 189 answers within two months, with additional answers still appearing. I have teamed up with a number of BPM researchers and practitioners to bring together our joint experience in a BPM workshop at the University in Liechtenstein in 2013, where we developed ten principles of good BPM, later published in Business Process Management Journal (vom Brocke et al., 2014). The paper, which has received considerable attention in academia, was ranked the journal’s most downloaded paper the month it was published. Slides on Slideshare that provide a brief summary of the paper have been accessed more than 3,000 times since they were first put online in March 2014. Given the importance of the topic–what makes for good BPM–and the positive response to the ten principles, I wrote this note with the co-authors of the original BPMJ paper to outline the ten principles and illustrate how to use them in practice. We invite all readers to engage in this discussion via any channel they find appropriate.

Conjugation of carcinogens by 6 class glutathione S-transferases : mechanisms and relevance to variations in human risk

Conjugation of chemicals with glutathione (GSH) can lead to decreased or increased toxicity. A genetic deficiency in the GSH S-transferase μ class gene M1 has been hypothesized to lead to greater risk of lung cancer in smokers. Recently a gene deletion polymorphism involving the human θ enzyme T1 has been described; the enzyme is present in erythrocytes and can be readily assayed. A rat θ class enzyme, 5-5, has structural and catalytic similarity and the protein was expressed in the Salmonella typhimurium tester strain TA1535. Expression of the cDNA vector increased the mutagenicity of ethylene dibromide and several methylene dihalides. Mutations resulting from the known GSH S-transferase substrate 1,2-epoxy-3-(4′nitrophenoxy)propane were decreased in the presence of the transferase. Expression of transferase 5-5 increased mutations when 1,2,3,4-diepoxybutane (butadiene diepoxide), 4-bromo-1,2-epoxybutane, or 1,3-dichloracetone were added. The latter compound is a model for the putative 1,2-dibromo-3-chloropropane oxidation product 1-bromo-3-chloroacetone. These genotoxicity and genotyping assays may be of use in further studies of the roles of GSH S-transferase θ enzymes in bioactivation and detoxication and any changes in risk due to polymorphism.

Identification of theta-class glutathione S-transferase in liver cytosol of the marmoset monkey

The presence of theta-class glutathione S-transferase (GST) in marmoset monkey liver cytosol was investigated. An anti-peptide antibody targeted against the C-terminus of rGSTT1 reacted with a single band in marmoset liver cytosol that corresponded to a molecular weight of 28 kDa. The intensity of the immunoreactive band was not affected by treatment of marmoset monkeys with 2,3,7,8-tetrachlorodibenzo-p-dioxin, phenobarbitone, rifampicin or clofibric acid. Similarly, activity towards methyl chloride (MC) was unaffected by these treatments. However, GST activity towards 1,2-epoxy3-(p- nitrophenoxy)-propane (EPNP) was increased in marmosets treated with phenobarbitone (2.6-fold) and rifampicin (2.6-fold), activity towards dichloromethane (DCM) was increased by 50% after treatment of marmosets with clofibric acid, and activity towards 1-chloro-2,4-dinitrobenzene (CDNB) was raised slightly (30-42% increases) after treatment with phenobarbitone, rifampicin or clofibric acid. Compared with humans, marmoset liver cytosol GST activity towards DCM was 18-fold higher, activity towards MC was 7 times higher and activity towards CDNB was 4 times higher. Further, EPNP activity was clearly detectable in marmoset liver cytosol samples, but was undetectable in human samples. Immunoreactive marmoset GST was partially purified by affinity chromatography using hexylglutathione-Sepharose and Orange A resin. The interaction of immunoreactive marmoset GST was similar to that found previously for rat and human GSTT1, suggesting that this protein is also a theta class GST. However, unlike rat GSTT1, the marmoset enzyme was not the major catalyst of EPNP conjugation. Instead, immunoreactivity was closely associated with activity towards MC. In conclusion, these results provide evidence for the presence of theta-class GST in the marmoset monkey orthologous to rGSTT1 and hGSTT1.

Model choice problems using approximate Bayesian computation with applications to pathogen transmission data sets

Analytically or computationally intractable likelihood functions can arise in complex statistical inferential problems making them inaccessible to standard Bayesian inferential methods. Approximate Bayesian computation (ABC) methods address such inferential problems by replacing direct likelihood evaluations with repeated sampling from the model. ABC methods have been predominantly applied to parameter estimation problems and less to model choice problems due to the added difficulty of handling multiple model spaces. The ABC algorithm proposed here addresses model choice problems by extending Fearnhead and Prangle (2012, Journal of the Royal Statistical Society, Series B 74, 1–28) where the posterior mean of the model parameters estimated through regression formed the summary statistics used in the discrepancy measure. An additional stepwise multinomial logistic regression is performed on the model indicator variable in the regression step and the estimated model probabilities are incorporated into the set of summary statistics for model choice purposes. A reversible jump Markov chain Monte Carlo step is also included in the algorithm to increase model diversity for thorough exploration of the model space. This algorithm was applied to a validating example to demonstrate the robustness of the algorithm across a wide range of true model probabilities. Its subsequent use in three pathogen transmission examples of varying complexity illustrates the utility of the algorithm in inferring preference of particular transmission models for the pathogens.

A pseudo-marginal sequential Monte Carlo algorithm for random effects models in Bayesian sequential design

A computationally efficient sequential Monte Carlo algorithm is proposed for the sequential design of experiments for the collection of block data described by mixed effects models. The difficulty in applying a sequential Monte Carlo algorithm in such settings is the need to evaluate the observed data likelihood, which is typically intractable for all but linear Gaussian models. To overcome this difficulty, we propose to unbiasedly estimate the likelihood, and perform inference and make decisions based on an exact-approximate algorithm. Two estimates are proposed: using Quasi Monte Carlo methods and using the Laplace approximation with importance sampling. Both of these approaches can be computationally expensive, so we propose exploiting parallel computational architectures to ensure designs can be derived in a timely manner. We also extend our approach to allow for model uncertainty. This research is motivated by important pharmacological studies related to the treatment of critically ill patients.

Music for the Pria Dewasa : changes and continuities in class and pop music genres

This paper presents Rolling Stone Indonesia (RSI) and places it in an historical context to tease out some changes and continuities in Indonesian middle-class politics since the beginning of the New Order. Some political scientists have claimed that class interests were at the core of the transition from Guided Democracy to the New Order, and popular music scholars generally assert that class underlies pop genre distinctions. But few have paid attention to how class and genre were written into Indonesian pop in the New Order period; Indonesian pop has a fascinating political history that has so far been overlooked. Placing RSI in historical perspective can reveal much about the print media’s classing of pop under New Order era political constraints, and about the ways these modes of classing may or may not have endured in the post-authoritarian, globalised and liberalised media environment.

Understanding Uncertainties in Non-Linear Population Trajectories: A Bayesian Semi-Parametric Hierarchical Approach to Large-Scale Surveys of Coral Cover

Recently, attempts to improve decision making in species management have focussed on uncertainties associated with modelling temporal fluctuations in populations. Reducing model uncertainty is challenging; while larger samples improve estimation of species trajectories and reduce statistical errors, they typically amplify variability in observed trajectories. In particular, traditional modelling approaches aimed at estimating population trajectories usually do not account well for nonlinearities and uncertainties associated with multi-scale observations characteristic of large spatio-temporal surveys. We present a Bayesian semi-parametric hierarchical model for simultaneously quantifying uncertainties associated with model structure and parameters, and scale-specific variability over time. We estimate uncertainty across a four-tiered spatial hierarchy of coral cover from the Great Barrier Reef. Coral variability is well described; however, our results show that, in the absence of additional model specifications, conclusions regarding coral trajectories become highly uncertain when considering multiple reefs, suggesting that management should focus more at the scale of individual reefs. The approach presented facilitates the description and estimation of population trajectories and associated uncertainties when variability cannot be attributed to specific causes and origins. We argue that our model can unlock value contained in large-scale datasets, provide guidance for understanding sources of uncertainty, and support better informed decision making

Combining opinions for use in Bayesian networks: A measurement error approach

Bayesian networks (BNs) are graphical probabilistic models used for reasoning under uncertainty. These models are becoming increasing popular in a range of fields including ecology, computational biology, medical diagnosis, and forensics. In most of these cases, the BNs are quantified using information from experts, or from user opinions. An interest therefore lies in the way in which multiple opinions can be represented and used in a BN. This paper proposes the use of a measurement error model to combine opinions for use in the quantification of a BN. The multiple opinions are treated as a realisation of measurement error and the model uses the posterior probabilities ascribed to each node in the BN which are computed from the prior information given by each expert. The proposed model addresses the issues associated with current methods of combining opinions such as the absence of a coherent probability model, the lack of the conditional independence structure of the BN being maintained, and the provision of only a point estimate for the consensus. The proposed model is applied an existing Bayesian Network and performed well when compared to existing methods of combining opinions.

Approximation of Bayesian predictive p-values with regression ABC

In the Bayesian framework a standard approach to model criticism is to compare some function of the observed data to a reference predictive distribution. The result of the comparison can be summarized in the form of a p-value, and it's well known that computation of some kinds of Bayesian predictive p-values can be challenging. The use of regression adjustment approximate Bayesian computation (ABC) methods is explored for this task. Two problems are considered. The first is the calibration of posterior predictive p-values so that they are uniformly distributed under some reference distribution for the data. Computation is difficult because the calibration process requires repeated approximation of the posterior for different data sets under the reference distribution. The second problem considered is approximation of distributions of prior predictive p-values for the purpose of choosing weakly informative priors in the case where the model checking statistic is expensive to compute. Here the computation is difficult because of the need to repeatedly sample from a prior predictive distribution for different values of a prior hyperparameter. In both these problems we argue that high accuracy in the computations is not required, which makes fast approximations such as regression adjustment ABC very useful. We illustrate our methods with several samples.

Contributions to Bayesian experimental design

This thesis progresses Bayesian experimental design by developing novel methodologies and extensions to existing algorithms. Through these advancements, this thesis provides solutions to several important and complex experimental design problems, many of which have applications in biology and medicine. This thesis consists of a series of published and submitted papers. In the first paper, we provide a comprehensive literature review on Bayesian design. In the second paper, we discuss methods which may be used to solve design problems in which one is interested in finding a large number of (near) optimal design points. The third paper presents methods for finding fully Bayesian experimental designs for nonlinear mixed effects models, and the fourth paper investigates methods to rapidly approximate the posterior distribution for use in Bayesian utility functions.

Pre-processing for approximate Bayesian computation in image analysis

Most of the existing algorithms for approximate Bayesian computation (ABC) assume that it is feasible to simulate pseudo-data from the model at each iteration. However, the computational cost of these simulations can be prohibitive for high dimensional data. An important example is the Potts model, which is commonly used in image analysis. Images encountered in real world applications can have millions of pixels, therefore scalability is a major concern. We apply ABC with a synthetic likelihood to the hidden Potts model with additive Gaussian noise. Using a pre-processing step, we fit a binding function to model the relationship between the model parameters and the synthetic likelihood parameters. Our numerical experiments demonstrate that the precomputed binding function dramatically improves the scalability of ABC, reducing the average runtime required for model fitting from 71 hours to only 7 minutes. We also illustrate the method by estimating the smoothing parameter for remotely sensed satellite imagery. Without precomputation, Bayesian inference is impractical for datasets of that scale.

Challenging the myth of an "epidemic" of common mental disorders : trends in the global prevalence of anxiety and depression between 1990 and 2010

Background Anxiety disorders and major depressive disorder (MDD) are common and disabling mental disorders. This paper aims to test the hypothesis that common mental disorders have become more prevalent over the past two decades. Methods We conducted a systematic review of prevalence, remission, duration, and excess mortality studies for anxiety disorders and MDD and then used a Bayesian meta-regression approach to estimate point prevalence for 1990, 2005, and 2010. We also conducted a post-hoc search for studies that used the General Health Questionnaire (GHQ) as a measure of psychological distress and tested for trends to present a qualitative comparison of study findings. Results This study found no evidence for an increased prevalence of anxiety disorders or MDD. While the crude number of cases increased by 36%, this was explained by population growth and changing age structures. Point prevalence of anxiety disorders was estimated at 3.8% (3.6-4.1%) in 1990 and 4.0% (3.7-4.2%) in 2010. The prevalence of MDD was unchanged at 4.4% in 1990 (4.2-4.7%) and 2010 (4.1-4.7%). However, 8 of the 11 GHQ studies found a significant increase in psychological distress over time. Conclusions The perceived "epidemic" of common mental disorders is most likely explained by the increasing numbers of affected patients driven by increasing population sizes. Additional factors that may explain this perception include the higher rates of psychological distress as measured using symptom checklists, greater public awareness, and the use of terms such as anxiety and depression in a context where they do not represent clinical disorders.