Matthew Denwood

Matthew Denwood

Associate Professor
‌Section for Animal Welfare and Disease Control
University of Copenhagen
Grønnegårdsvej 8
1870 Frederiksberg C
Denmark

Link to my page at the University of Copenhagen

Research Interests

Research Interests

My main research interests are in the development and application of computationally intensive statistical techniques to complex biological problems, particularly using techniques such as Bayesian Markov chain Monte Carlo (MCMC), and the use of stochastic modelling as a tool to understand disease transmission. I also have a variety of ongoing research collaborations in a broad range of epidemiological contexts, including modelling parasite distributions, disease surveillance, appropriate handling of count data such as faecal egg count reduction test (FECRT) data, and in analysis of a wide range of clinical and ecological datasets.

Academic History

Academic History

Associate Professor in Quantitative Veterinary Epidemiology and Biostatistics, University of Copenhagn, 2014 - current

Lecturer in Production Animal Health, University of Glasgow, 2012-2014.

Post doctoral research assistant - Why some hosts have high parasite burdens, and the implications for the design of sustainable control strategies, University of Glasgow, 2011-2012.

Research fellow with EPIC - The Scottish Government's Centre of Excellence in Epidemiology, Population Health and Disease Control, University of Glasgow, 2010-2011.

PhD, University of Glasgow - 'Use of Bayesian MCMC to analyse the distribution of faecal egg counts' supervised by Sandy love, Giles Innocent and Stuart Reid, 2006-2010.  The thesis can be downloaded from here.

BVMS, University of Glasgow, 2001-2006.

Software Links

Software Links

I have created the R package runjags which contains a suite of functions to allow the R statistical programming language to interact with the MCMC tool 'Just Another Gibbs Sampler' (JAGS). Any user specified model can be run in JAGS, with several functions intended to automate the process of setting up and monitoring the simulation for convergence, using multiple cores to speed up computations using parallel processing, and returning the simulation results as an R object.  The R package bayescount contains functions to analyse count datasets, including feacal egg count reduction tests, using a variety of (zero-inflated) distributions implemented using MCMC. This package also contains functions to perform power calculations for both FEC and FECRT studies. Both packages, along with the software needed to run them, are freely available from the links provided, and have been used by researchers internationally and cited as part of peer reviewed publications.