A5100649922- Statistical Methods and Inference
- Bayesian Methods and Mixture Models
- Statistical Methods and Bayesian Inference
- Gene expression and cancer classification
- Genetic Associations and Epidemiology
- Bayesian Modeling and Causal Inference
- Gaussian Processes and Bayesian Inference
- Generative Adversarial Networks and Image Synthesis
- Genetic and phenotypic traits in livestock
- Markov Chains and Monte Carlo Methods
- Advanced Causal Inference Techniques
- Bioinformatics and Genomic Networks
- Adversarial Robustness in Machine Learning
- Advanced Statistical Methods and Models
- Genomic variations and chromosomal abnormalities
- Genetic Mapping and Diversity in Plants and Animals
- COVID-19 epidemiological studies
- Statistical Methods in Clinical Trials
- Music and Audio Processing
- Neural Networks and Applications
- Malaria Research and Control
- Single-cell and spatial transcriptomics
- Anomaly Detection Techniques and Applications
- Artificial Intelligence in Healthcare and Education
- Explainable Artificial Intelligence (XAI)
University of Oxford
2016-2025
The Alan Turing Institute
2019-2024
Turing Institute
2020-2023
Mary Lyon Centre at MRC Harwell
2008-2022
Health Data Research UK
2020-2022
Royal Statistical Society
2021
Open Data Institute
2018-2021
Centre for Human Genetics
2009-2020
University of Warwick
2020
Cambridge Military Hospital
2020
In the past ten years there has been a dramatic increase of interest in Bayesian analysis finite mixture models. This is primarily because emergence Markov chain Monte Carlo (MCMC) methods. While MCMC provides convenient way to draw inference from complicated statistical models, are many, perhaps underappreciated, problems associated with mixtures. The mainly caused by nonidentifiability components under symmetric priors, which leads so-called label switching output. means that ergodic...
Array-based technologies have been used to detect chromosomal copy number changes (aneuploidies) in the human genome.Recent studies identified numerous variants (CNV ) and some are common polymorphisms that may contribute disease susceptibility.We developed, experimentally validated, a novel computational framework (QuantiSNP) for detecting regions of variation from BeadArray TM SNP genotyping data using an Objective Bayes Hidden-Markov Model (OB-HMM).Objective measures set certain...
BackgroundThe medical, societal, and economic impact of the coronavirus disease 2019 (COVID-19) pandemic has unknown effects on overall population mortality. Previous models mortality are based death over days among infected people, nearly all whom thus far have underlying conditions. Models not incorporated information high-risk conditions or their longer-term baseline (pre-COVID-19) We estimated excess number deaths 1 year under different COVID-19 incidence scenarios varying levels...
In this paper we discuss auxiliary variable approaches to Bayesian binary and multinomial regression. These are ideally suited automated Markov chain Monte Carlo simulation. the first part describe a simple technique using joint updating that improves performance of conventional probit regression algorithm. second methods for inference in logistic regression, including covariate set uncertainty. Finally, show how method is easily extended models. All algorithms fully automatic with no user...
Machine learning, artificial intelligence, and other modern statistical methods are providing new opportunities to operationalise previously untapped rapidly growing sources of data for patient benefit. Despite much promising research currently being undertaken, particularly in imaging, the literature as a whole lacks transparency, clear reporting facilitate replicability, exploration potential ethical concerns, demonstrations effectiveness. Among many reasons why these problems exist, one...
We present a case study on the utility of graphics cards to perform massively parallel simulation advanced Monte Carlo methods. Graphics cards, containing multiple Processing Units (GPUs), are self-contained computational devices that can be housed in conventional desktop and laptop computers thought as prototypes next generation many-core processors. For certain classes population-based algorithms they offer simulation, with added advantage over distributed multicore processors cheap,...
We propose a framework for general Bayesian inference. argue that valid update of prior belief distribution to posterior can be made parameters which are connected observations through loss function rather than the traditional likelihood function, is recovered under special case using self information loss. Modern application areas make it increasingly challenging Bayesians attempt model true data generating mechanism. Moreover, when object interest low dimensional, such as mean or median,...
Drawing from real-life scenarios and insights shared at the RAISE (Responsible AI for Social Ethical Healthcare) conference, we highlight critical need in health care (AIH) to primarily benefit patients address current shortcomings systems such as medical errors access disparities. The embodying a sense of responsibility urgency, emphasized that AIH should enhance patient care, support professionals, be accessible safe all. discussions revolved around immediate actions leaders, adopting...
The authors describe the development of a four-dimensional atlas and reference system that includes both macroscopic microscopic information on structure function human brain in persons between ages 18 90 years. Given presumed large but previously unquantified degree structural functional variance among normal population, basis for this is probabilistic. Through efforts International Consortium Brain Mapping (ICBM), 7,000 subjects will be included initial phase database development. For each...
AbstractIn many problems in geostatistics the response variable of interest is strongly related to underlying geology spatial location. In these situations there often little correlation responses found different rock strata, so covariance structure shows sharp changes at boundaries types. Conventional stationary and nonstationary methods are inappropriate, because they typically assume that between points a smooth function distance. this article we propose generic method for analysis data...
Malaria represents one of the major worldwide challenges to public health. A recent breakthrough in study disease follows annotation genome malaria parasite Plasmodium falciparum and mosquito vector (an organism that spreads an infectious disease)Anopheles. Of particular interest is molecular biology underlying immune response system Anopheles, which actively fights against infection. This article reports a statistical analysis gene expression time profiles from mosquitoes have been infected...
Upper- and lower-body fat depots exhibit opposing associations with obesity-related metabolic disease. We defined the relationship between DEXA-quantified diabetes/cardiovascular risk factors in a healthy population-based cohort (n = 3,399). Gynoid mass correlated negatively insulin resistance after total adjustment, whereas opposite was seen for abdominal fat. Paired transcriptomic analysis of gluteal subcutaneous adipose tissue (GSAT) (ASAT) performed across BMI spectrum 49; 21.4–45.5...
Highly recombinant populations derived from inbred lines, such as advanced intercross lines and heterogeneous stocks, can be used to map loci far more accurately than is possible with standard intercrosses. However, the varying degrees of relatedness that exist between individuals complicate analysis, potentially leading many false positive signals. We describe a method deal these problems does not require pedigree information accounts for model uncertainty through averaging. In our method,...