Rana Moyeed - Academia.edu (original) (raw)

Papers by Rana Moyeed

Research paper thumbnail of Additional file 2: of Gene expression meta-analysis of Parkinson's disease and its relationship with Alzheimer's disease

DEGs identified in this study. A full list of the 1046 DEGs identified in this meta-analysis in a... more DEGs identified in this study. A full list of the 1046 DEGs identified in this meta-analysis in an Excel file. (XLSX 120 kb)

Research paper thumbnail of Additional file 1: of Gene expression meta-analysis of Parkinson's disease and its relationship with Alzheimer's disease

Table S1. Information about each study used in our meta-analysis after removal of outlier samples... more Table S1. Information about each study used in our meta-analysis after removal of outlier samples. Table S2. Differentially expressed genes identified in our meta-analysis that have been identified as PD risk genes in a recent GWAS meta-analysis [33]. Table S3. IPA canonical pathway analysis for significant pathways identified using all PD DEGs, included with the information for pathways shared with those identified as significant using all AD DEGs. Table S4. IPA canonical pathway analysis for significant pathways identified using down-regulated PD DEGs. Table S5. IPA upstream regulator analysis for up and down regulated PD DEGs analysed separately. Table S6. Top 10 hubs found in the protein-protein interaction network (PPIN) analysis subnetwork created using the top 30 PD DEGs. Table S7. The direction of differential expression between the common DEGs found between AD and PD. Figure S1. Selecting filtering threshold for microarray data. The percentage of studies called absent in a ...

Research paper thumbnail of Bayesian statistics in the design and analysis of cluster randomised controlled trials and their reporting quality: a methodological systematic review

Systematic Reviews, 2021

Background In a cluster randomised controlled trial (CRCT), randomisation units are “clusters” su... more Background In a cluster randomised controlled trial (CRCT), randomisation units are “clusters” such as schools or GP practices. This has methodological implications for study design and statistical analysis, since clustering often leads to correlation between observations which, if not accounted for, can lead to spurious conclusions of efficacy/effectiveness. Bayesian methodology offers a flexible, intuitive framework to deal with such issues, but its use within CRCT design and analysis appears limited. This review aims to explore and quantify the use of Bayesian methodology in the design and analysis of CRCTs, and appraise the quality of reporting against CONSORT guidelines. Methods We sought to identify all reported/published CRCTs that incorporated Bayesian methodology and papers reporting development of new Bayesian methodology in this context, without restriction on publication date or location. We searched Medline and Embase and the Cochrane Central Register of Controlled Tria...

Research paper thumbnail of Bayesian modelling of ultra high-frequency financial data

International Journal of Statistics and Economics, 2011

The availability of ultra high-frequency (UHF) financial data on transactions has revolutionised ... more The availability of ultra high-frequency (UHF) financial data on transactions has revolutionised statistical modelling techniques in finance. The unique characteristics of such data, e.g. discrete structure of price change and unequally spaced time intervals have introduced new challenges to statistical studies. In this study, we develop a Bayesian framework for modelling integer-valued variables to capture the behaviour of price change. We propose the application of the zero inflated Poisson difference (ZPD) distribution and assess the effect of covariates. We apply our model to a set of FTSE100 index changes and obtain the predictive distribution of the index change. We then use the deviance information criterion for the purpose of model comparison. Finally, based on the probability integral transform, modified for the case of integer-valued variables, we show that our model is capable of explaining well the observed distribution of price change.

Research paper thumbnail of Stochastic approximation of the MLE for a spatial point pattern

Physica D: Nonlinear Phenomena, 1989

An iterative stochastic approximation to the maximum Ukelihood estimate is developed for the Stra... more An iterative stochastic approximation to the maximum Ukelihood estimate is developed for the Strauss point process. We modify existing theorems to show that the approximation is consistent and asymptotically normal. It performs well in numerical tests.

Research paper thumbnail of Bayesian hierarchical models for linear networks

Journal of Applied Statistics

Research paper thumbnail of Genetic networks in Parkinson’s and Alzheimer’s disease

Aging

Parkinson's disease (PD) and Alzheimer's disease (AD) are the most common neurodegenerative disea... more Parkinson's disease (PD) and Alzheimer's disease (AD) are the most common neurodegenerative diseases and there is increasing evidence that they share common physiological and pathological links. Here we have conducted the largest network analysis of PD and AD based on their gene expressions in blood to date. We identified modules that were not preserved between disease and healthy control (HC) networks, and important hub genes and transcription factors (TFs) in these modules. We highlighted that the PD module not preserved in HCs was associated with insulin resistance, and HDAC6 was identified as a hub gene in this module which may have the role of influencing tau phosphorylation and autophagic flux in neurodegenerative disease. The AD module associated with regulation of lipolysis in adipocytes and neuroactive ligand-receptor interaction was not preserved in healthy and mild cognitive impairment networks and the key hubs TRPC5 and BRAP identified as potential targets for therapeutic treatments of AD. Our study demonstrated that PD and AD share common disrupted genetics and identified novel pathways, hub genes and TFs that may be new areas for mechanistic study and important targets in both diseases.

Research paper thumbnail of Gene expression meta-analysis of Parkinson’s disease and its relationship with Alzheimer’s disease

Molecular Brain

Parkinson's disease (PD) and Alzheimer's disease (AD) are the most common neurodegenerative disea... more Parkinson's disease (PD) and Alzheimer's disease (AD) are the most common neurodegenerative diseases and have been suggested to share common pathological and physiological links. Understanding the cross-talk between them could reveal potentials for the development of new strategies for early diagnosis and therapeutic intervention thus improving the quality of life of those affected. Here we have conducted a novel meta-analysis to identify differentially expressed genes (DEGs) in PD microarray datasets comprising 69 PD and 57 control brain samples which is the biggest cohort for such studies to date. Using identified DEGs, we performed pathway, upstream and protein-protein interaction analysis. We identified 1046 DEGs, of which a majority (739/1046) were downregulated in PD. YWHAZ and other genes coding 14-3-3 proteins are identified as important DEGs in signaling pathways and in protein-protein interaction networks (PPIN). Perturbed pathways also include mitochondrial dysfunction and oxidative stress. There was a significant overlap in DEGs between PD and AD, and over 99% of these were differentially expressed in the same up or down direction across the diseases. REST was identified as an upstream regulator in both diseases. Our study demonstrates that PD and AD share significant common DEGs and pathways, and identifies novel genes, pathways and upstream regulators which may be important targets for therapy in both diseases.

Research paper thumbnail of Canonical Correlation Analysis When the Data are Curves

Journal of the Royal Statistical Society: Series B (Methodological)

ABSTRACT It is not immediately straightforward to extend canonical correlation analysis to the co... more ABSTRACT It is not immediately straightforward to extend canonical correlation analysis to the context of functional data analysis, where the data are themselves curves or functions. The obvious approach breaks down, and it is necessary to use a method involving smoothing in some way. Such a method is introduced and discussed with reference to a data set on human gait. The breakdown of the unsmoothed method is illustrated in a practical context and is demonstrated theoretically. A consistency theorem for the smoothed method is proved.

Research paper thumbnail of Linking genotoxicity and cytotoxicity with membrane fluidity: A comparative study in ovarian cancer cell lines following exposure to auranofin

Mutation Research/Genetic Toxicology and Environmental Mutagenesis, 2016

Auranofin, an organogold compound classified as an anti-rheumatic agent is under phase 2 clinical... more Auranofin, an organogold compound classified as an anti-rheumatic agent is under phase 2 clinical trials for re-purposing to treat recurrent epithelial ovarian cancer. We have reported earlier that Breast cancer 1, early onset (BRCA1) mutant ovarian cancer cells exhibit increased sensitivity to auranofin. BRCA1 is a DNA repair protein whose functional status is critical in the prognosis of ovarian cancer. Apart from DNA repair capability of cancer cells, membrane fluidity is also implicated in modulating resistance to chemotherapeutics. We report here that membrane fluidity influences the sensitivity of ovarian cancer cell lines (OVCAR5 and IGROV1) to auranofin. Electron spin resonance (ESR) analysis revealed a more fluidized membrane in IGROV1 compared to OVCAR5. Interestingly, IGROV1 cells were more sensitive to auranofin induced cytotoxicity than OVCAR5. In comparison to OVCAR5, IGROV1 cells also exhibited an increased number of DNA double strand breaks (DSBs) upon auranofin treatment as assessed by 53BP1 immunostaining. Furthermore, correlation analysis demonstrated a strong positive correlation (r = 0.856) between membrane fluidity and auranofin sensitivity in these cell lines. Auranofin-treated IGROV1 cells were also exhibited increased cellular oxidation and apoptosis. Anti-oxidant, N-acetyl cysteine (NAC) inhibited the cellular oxidation and apoptosis in auranofin-treated ovarian cancer cells suggesting reactive oxygen species (ROS) mediates the anti-cancer properties of auranofin. Overall, our study suggests that auranofin mediates its cytotoxicity via ROS production in ovarian cancer cells which correlates positively with membrane fluidity.

Research paper thumbnail of Model-based geostatistics

J Roy Stat Soc Ser C Appl, 2002

Research paper thumbnail of An application of the generalized Poisson difference distribution to the Bayesian modelling of football scores

Statistica Neerlandica, 2016

The analysis of sports data, in particular football match outcomes, has always produced an immens... more The analysis of sports data, in particular football match outcomes, has always produced an immense interest among the statisticians. In this paper, we adopt the generalised Poisson difference distribution (GPDD) to model the goal difference of football matches. We discuss the advantages of the proposed model over the Poisson difference (PD) model which was also used for the same purpose. The GPDD model, like the PD model, is based on the goal difference in each game which allows us to account for the correlation without explicitly modelling it. The main advantage of the GPDD model is its flexibility in the tails by considering shorter as well as longer tails than the PD distribution. We carry out the analysis in a Bayesian framework in order to incorporate external information, such as historical knowledge or data, through the prior distributions. We model both the mean and the variance of the goal difference and show that such a model performs considerably better than a model with a fixed variance. Finally, the proposed model is fitted to the 2012-13 Italian Serie A football data and various model diagnostics are carried out to evaluate the performance of the model.

Research paper thumbnail of Bayesian nonparametric quantile regression using splines

Computational Statistics Data Analysis, Apr 1, 2010

A new technique based on Bayesian quantile regression that models the dependence of a quantile of... more A new technique based on Bayesian quantile regression that models the dependence of a quantile of one variable on the values of another using a natural cubic spline is presented. Inference is based on the posterior density of the spline and an associated smoothing parameter and ...

Research paper thumbnail of Model-based geostatistics (with discussion)

Research paper thumbnail of Linear and nonlinear kriging methods: Tools for monitoring soil pollution

Research paper thumbnail of Parameter Uncertainty in Spatial Prediction: Checking its Importance by Cross-Validating the Wolfcamp and Rongelap Data Sets

Quantitative Geology and Geostatistics, 2001

Traditional geostatistical prediction techniques assume that the covariance structure of the data... more Traditional geostatistical prediction techniques assume that the covariance structure of the data is known. In practice, the covariance must be estimated from data, and the estimate is used for computing predictions. The additional parameter uncertainty about the covariance structure is therefore not properly taken into account by customary plug-in kriging methods. Bayesian (Kitanidis, 1986; Handcock and Stein, 1993) and model-based kriging (Diggle et al., 1998) naturally incorporate parameter uncertainty into the predictions. In this study, we compare model-based and plug-in kriging methods, using two sets of data: the pressure head of the Wolfcamp aquifer and the 173caesium concentration in the ground of Rongelap Island. We used the precision of the predictions and the success in modelling the prediction uncertainty as criteria to rank the methods. The main results were: (i) plug-in kriging methods were as precise as model-based kriging, (ii) linear kriging successfully modelled prediction uncertainty, provided the marginal distribution was close to normal and the variogram was unbiasedly estimated for non-stationary data, (iii) model-based kriging failed to model the 137Cs concentration accurately. Given these results and our experiences from an empirical comparison of non-linear kriging methods (Moyeed and Papritz, 2000; Papritz and Moyeed, 1999), we would suggest that the question of parameter uncertainty be looked into more closely.

Research paper thumbnail of Rates of Convergence in Semi-Parametric Modelling of Longitudinal Data

Australian Journal of Statistics, 1994

Summary We consider the problem of semi-parametric regression modelling when the data consist of ... more Summary We consider the problem of semi-parametric regression modelling when the data consist of a collection of short time series for which measurements within series are correlated. The objective is to estimate a regression function of the form E [Y (t)| x]= x'ß+ μ ( ...

Research paper thumbnail of Spline Smoother as a Dynamic Linear Model

Australian Journal of Statistics, 1995

ABSTRACT The paper shows how a finite dimensional representation of a cubic smoothing spline can ... more ABSTRACT The paper shows how a finite dimensional representation of a cubic smoothing spline can be put in the framework of a dynamic linear model. The formulation provides an updating scheme when observations do not occur sequentially in time or space.

Research paper thumbnail of Spatial variation of Anopheles-transmitted Wuchereria bancrofti and Plasmodium falciparum infection densities in Papua New Guinea

Filaria journal, Jan 14, 2003

The spatial variation of Wuchereria bancrofti and Plasmodium falciparum infection densities was m... more The spatial variation of Wuchereria bancrofti and Plasmodium falciparum infection densities was measured in a rural area of Papua New Guinea where they share anopheline vectors. The spatial correlation of W. bancrofti was found to reduce by half over an estimated distance of 1.7 km, much smaller than the 50 km grid used by the World Health Organization rapid mapping method. For P. falciparum, negligible spatial correlation was found. After mass treatment with anti-filarial drugs, there was negligible correlation between the changes in the densities of the two parasites.

Research paper thumbnail of Stochastic approximation of the MLE for a spatial point pattern

ABSTRACT An iterative stochastic approximation to the maximum likelihood estimate is developed fo... more ABSTRACT An iterative stochastic approximation to the maximum likelihood estimate is developed for the Strauss point process. We modify existing theorems to show that the approximation is consistent and asymptotically normal. It performs well in numerical tests.

Research paper thumbnail of Additional file 2: of Gene expression meta-analysis of Parkinson's disease and its relationship with Alzheimer's disease

DEGs identified in this study. A full list of the 1046 DEGs identified in this meta-analysis in a... more DEGs identified in this study. A full list of the 1046 DEGs identified in this meta-analysis in an Excel file. (XLSX 120 kb)

Research paper thumbnail of Additional file 1: of Gene expression meta-analysis of Parkinson's disease and its relationship with Alzheimer's disease

Table S1. Information about each study used in our meta-analysis after removal of outlier samples... more Table S1. Information about each study used in our meta-analysis after removal of outlier samples. Table S2. Differentially expressed genes identified in our meta-analysis that have been identified as PD risk genes in a recent GWAS meta-analysis [33]. Table S3. IPA canonical pathway analysis for significant pathways identified using all PD DEGs, included with the information for pathways shared with those identified as significant using all AD DEGs. Table S4. IPA canonical pathway analysis for significant pathways identified using down-regulated PD DEGs. Table S5. IPA upstream regulator analysis for up and down regulated PD DEGs analysed separately. Table S6. Top 10 hubs found in the protein-protein interaction network (PPIN) analysis subnetwork created using the top 30 PD DEGs. Table S7. The direction of differential expression between the common DEGs found between AD and PD. Figure S1. Selecting filtering threshold for microarray data. The percentage of studies called absent in a ...

Research paper thumbnail of Bayesian statistics in the design and analysis of cluster randomised controlled trials and their reporting quality: a methodological systematic review

Systematic Reviews, 2021

Background In a cluster randomised controlled trial (CRCT), randomisation units are “clusters” su... more Background In a cluster randomised controlled trial (CRCT), randomisation units are “clusters” such as schools or GP practices. This has methodological implications for study design and statistical analysis, since clustering often leads to correlation between observations which, if not accounted for, can lead to spurious conclusions of efficacy/effectiveness. Bayesian methodology offers a flexible, intuitive framework to deal with such issues, but its use within CRCT design and analysis appears limited. This review aims to explore and quantify the use of Bayesian methodology in the design and analysis of CRCTs, and appraise the quality of reporting against CONSORT guidelines. Methods We sought to identify all reported/published CRCTs that incorporated Bayesian methodology and papers reporting development of new Bayesian methodology in this context, without restriction on publication date or location. We searched Medline and Embase and the Cochrane Central Register of Controlled Tria...

Research paper thumbnail of Bayesian modelling of ultra high-frequency financial data

International Journal of Statistics and Economics, 2011

The availability of ultra high-frequency (UHF) financial data on transactions has revolutionised ... more The availability of ultra high-frequency (UHF) financial data on transactions has revolutionised statistical modelling techniques in finance. The unique characteristics of such data, e.g. discrete structure of price change and unequally spaced time intervals have introduced new challenges to statistical studies. In this study, we develop a Bayesian framework for modelling integer-valued variables to capture the behaviour of price change. We propose the application of the zero inflated Poisson difference (ZPD) distribution and assess the effect of covariates. We apply our model to a set of FTSE100 index changes and obtain the predictive distribution of the index change. We then use the deviance information criterion for the purpose of model comparison. Finally, based on the probability integral transform, modified for the case of integer-valued variables, we show that our model is capable of explaining well the observed distribution of price change.

Research paper thumbnail of Stochastic approximation of the MLE for a spatial point pattern

Physica D: Nonlinear Phenomena, 1989

An iterative stochastic approximation to the maximum Ukelihood estimate is developed for the Stra... more An iterative stochastic approximation to the maximum Ukelihood estimate is developed for the Strauss point process. We modify existing theorems to show that the approximation is consistent and asymptotically normal. It performs well in numerical tests.

Research paper thumbnail of Bayesian hierarchical models for linear networks

Journal of Applied Statistics

Research paper thumbnail of Genetic networks in Parkinson’s and Alzheimer’s disease

Aging

Parkinson's disease (PD) and Alzheimer's disease (AD) are the most common neurodegenerative disea... more Parkinson's disease (PD) and Alzheimer's disease (AD) are the most common neurodegenerative diseases and there is increasing evidence that they share common physiological and pathological links. Here we have conducted the largest network analysis of PD and AD based on their gene expressions in blood to date. We identified modules that were not preserved between disease and healthy control (HC) networks, and important hub genes and transcription factors (TFs) in these modules. We highlighted that the PD module not preserved in HCs was associated with insulin resistance, and HDAC6 was identified as a hub gene in this module which may have the role of influencing tau phosphorylation and autophagic flux in neurodegenerative disease. The AD module associated with regulation of lipolysis in adipocytes and neuroactive ligand-receptor interaction was not preserved in healthy and mild cognitive impairment networks and the key hubs TRPC5 and BRAP identified as potential targets for therapeutic treatments of AD. Our study demonstrated that PD and AD share common disrupted genetics and identified novel pathways, hub genes and TFs that may be new areas for mechanistic study and important targets in both diseases.

Research paper thumbnail of Gene expression meta-analysis of Parkinson’s disease and its relationship with Alzheimer’s disease

Molecular Brain

Parkinson's disease (PD) and Alzheimer's disease (AD) are the most common neurodegenerative disea... more Parkinson's disease (PD) and Alzheimer's disease (AD) are the most common neurodegenerative diseases and have been suggested to share common pathological and physiological links. Understanding the cross-talk between them could reveal potentials for the development of new strategies for early diagnosis and therapeutic intervention thus improving the quality of life of those affected. Here we have conducted a novel meta-analysis to identify differentially expressed genes (DEGs) in PD microarray datasets comprising 69 PD and 57 control brain samples which is the biggest cohort for such studies to date. Using identified DEGs, we performed pathway, upstream and protein-protein interaction analysis. We identified 1046 DEGs, of which a majority (739/1046) were downregulated in PD. YWHAZ and other genes coding 14-3-3 proteins are identified as important DEGs in signaling pathways and in protein-protein interaction networks (PPIN). Perturbed pathways also include mitochondrial dysfunction and oxidative stress. There was a significant overlap in DEGs between PD and AD, and over 99% of these were differentially expressed in the same up or down direction across the diseases. REST was identified as an upstream regulator in both diseases. Our study demonstrates that PD and AD share significant common DEGs and pathways, and identifies novel genes, pathways and upstream regulators which may be important targets for therapy in both diseases.

Research paper thumbnail of Canonical Correlation Analysis When the Data are Curves

Journal of the Royal Statistical Society: Series B (Methodological)

ABSTRACT It is not immediately straightforward to extend canonical correlation analysis to the co... more ABSTRACT It is not immediately straightforward to extend canonical correlation analysis to the context of functional data analysis, where the data are themselves curves or functions. The obvious approach breaks down, and it is necessary to use a method involving smoothing in some way. Such a method is introduced and discussed with reference to a data set on human gait. The breakdown of the unsmoothed method is illustrated in a practical context and is demonstrated theoretically. A consistency theorem for the smoothed method is proved.

Research paper thumbnail of Linking genotoxicity and cytotoxicity with membrane fluidity: A comparative study in ovarian cancer cell lines following exposure to auranofin

Mutation Research/Genetic Toxicology and Environmental Mutagenesis, 2016

Auranofin, an organogold compound classified as an anti-rheumatic agent is under phase 2 clinical... more Auranofin, an organogold compound classified as an anti-rheumatic agent is under phase 2 clinical trials for re-purposing to treat recurrent epithelial ovarian cancer. We have reported earlier that Breast cancer 1, early onset (BRCA1) mutant ovarian cancer cells exhibit increased sensitivity to auranofin. BRCA1 is a DNA repair protein whose functional status is critical in the prognosis of ovarian cancer. Apart from DNA repair capability of cancer cells, membrane fluidity is also implicated in modulating resistance to chemotherapeutics. We report here that membrane fluidity influences the sensitivity of ovarian cancer cell lines (OVCAR5 and IGROV1) to auranofin. Electron spin resonance (ESR) analysis revealed a more fluidized membrane in IGROV1 compared to OVCAR5. Interestingly, IGROV1 cells were more sensitive to auranofin induced cytotoxicity than OVCAR5. In comparison to OVCAR5, IGROV1 cells also exhibited an increased number of DNA double strand breaks (DSBs) upon auranofin treatment as assessed by 53BP1 immunostaining. Furthermore, correlation analysis demonstrated a strong positive correlation (r = 0.856) between membrane fluidity and auranofin sensitivity in these cell lines. Auranofin-treated IGROV1 cells were also exhibited increased cellular oxidation and apoptosis. Anti-oxidant, N-acetyl cysteine (NAC) inhibited the cellular oxidation and apoptosis in auranofin-treated ovarian cancer cells suggesting reactive oxygen species (ROS) mediates the anti-cancer properties of auranofin. Overall, our study suggests that auranofin mediates its cytotoxicity via ROS production in ovarian cancer cells which correlates positively with membrane fluidity.

Research paper thumbnail of Model-based geostatistics

J Roy Stat Soc Ser C Appl, 2002

Research paper thumbnail of An application of the generalized Poisson difference distribution to the Bayesian modelling of football scores

Statistica Neerlandica, 2016

The analysis of sports data, in particular football match outcomes, has always produced an immens... more The analysis of sports data, in particular football match outcomes, has always produced an immense interest among the statisticians. In this paper, we adopt the generalised Poisson difference distribution (GPDD) to model the goal difference of football matches. We discuss the advantages of the proposed model over the Poisson difference (PD) model which was also used for the same purpose. The GPDD model, like the PD model, is based on the goal difference in each game which allows us to account for the correlation without explicitly modelling it. The main advantage of the GPDD model is its flexibility in the tails by considering shorter as well as longer tails than the PD distribution. We carry out the analysis in a Bayesian framework in order to incorporate external information, such as historical knowledge or data, through the prior distributions. We model both the mean and the variance of the goal difference and show that such a model performs considerably better than a model with a fixed variance. Finally, the proposed model is fitted to the 2012-13 Italian Serie A football data and various model diagnostics are carried out to evaluate the performance of the model.

Research paper thumbnail of Bayesian nonparametric quantile regression using splines

Computational Statistics Data Analysis, Apr 1, 2010

A new technique based on Bayesian quantile regression that models the dependence of a quantile of... more A new technique based on Bayesian quantile regression that models the dependence of a quantile of one variable on the values of another using a natural cubic spline is presented. Inference is based on the posterior density of the spline and an associated smoothing parameter and ...

Research paper thumbnail of Model-based geostatistics (with discussion)

Research paper thumbnail of Linear and nonlinear kriging methods: Tools for monitoring soil pollution

Research paper thumbnail of Parameter Uncertainty in Spatial Prediction: Checking its Importance by Cross-Validating the Wolfcamp and Rongelap Data Sets

Quantitative Geology and Geostatistics, 2001

Traditional geostatistical prediction techniques assume that the covariance structure of the data... more Traditional geostatistical prediction techniques assume that the covariance structure of the data is known. In practice, the covariance must be estimated from data, and the estimate is used for computing predictions. The additional parameter uncertainty about the covariance structure is therefore not properly taken into account by customary plug-in kriging methods. Bayesian (Kitanidis, 1986; Handcock and Stein, 1993) and model-based kriging (Diggle et al., 1998) naturally incorporate parameter uncertainty into the predictions. In this study, we compare model-based and plug-in kriging methods, using two sets of data: the pressure head of the Wolfcamp aquifer and the 173caesium concentration in the ground of Rongelap Island. We used the precision of the predictions and the success in modelling the prediction uncertainty as criteria to rank the methods. The main results were: (i) plug-in kriging methods were as precise as model-based kriging, (ii) linear kriging successfully modelled prediction uncertainty, provided the marginal distribution was close to normal and the variogram was unbiasedly estimated for non-stationary data, (iii) model-based kriging failed to model the 137Cs concentration accurately. Given these results and our experiences from an empirical comparison of non-linear kriging methods (Moyeed and Papritz, 2000; Papritz and Moyeed, 1999), we would suggest that the question of parameter uncertainty be looked into more closely.

Research paper thumbnail of Rates of Convergence in Semi-Parametric Modelling of Longitudinal Data

Australian Journal of Statistics, 1994

Summary We consider the problem of semi-parametric regression modelling when the data consist of ... more Summary We consider the problem of semi-parametric regression modelling when the data consist of a collection of short time series for which measurements within series are correlated. The objective is to estimate a regression function of the form E [Y (t)| x]= x'ß+ μ ( ...

Research paper thumbnail of Spline Smoother as a Dynamic Linear Model

Australian Journal of Statistics, 1995

ABSTRACT The paper shows how a finite dimensional representation of a cubic smoothing spline can ... more ABSTRACT The paper shows how a finite dimensional representation of a cubic smoothing spline can be put in the framework of a dynamic linear model. The formulation provides an updating scheme when observations do not occur sequentially in time or space.

Research paper thumbnail of Spatial variation of Anopheles-transmitted Wuchereria bancrofti and Plasmodium falciparum infection densities in Papua New Guinea

Filaria journal, Jan 14, 2003

The spatial variation of Wuchereria bancrofti and Plasmodium falciparum infection densities was m... more The spatial variation of Wuchereria bancrofti and Plasmodium falciparum infection densities was measured in a rural area of Papua New Guinea where they share anopheline vectors. The spatial correlation of W. bancrofti was found to reduce by half over an estimated distance of 1.7 km, much smaller than the 50 km grid used by the World Health Organization rapid mapping method. For P. falciparum, negligible spatial correlation was found. After mass treatment with anti-filarial drugs, there was negligible correlation between the changes in the densities of the two parasites.

Research paper thumbnail of Stochastic approximation of the MLE for a spatial point pattern

ABSTRACT An iterative stochastic approximation to the maximum likelihood estimate is developed fo... more ABSTRACT An iterative stochastic approximation to the maximum likelihood estimate is developed for the Strauss point process. We modify existing theorems to show that the approximation is consistent and asymptotically normal. It performs well in numerical tests.