Joe O'Reilly

epidemiology - statistical inference - phylogenetics - data science

About Me


Originally from East Yorkshire, but currently living in Bristol (I work remotely at The University of Edinburgh), my work involves the application of a range of statistical methods to gain unique insights into complex, large, datasets. I am particularly interested in the use of Bayesian approaches to the estimation of model parameters and their associated uncertainty. I currently employ such methods to electronic healthcare record data and clinical trial data in my role as a postdoctoral epidemiologist.

My Work


I am currently employed as a postdoctoral biostatistician/epidemiologist at The University of Edinburgh, based at The Institute of Genetics and Cancer. The research I am involved with uses national level electronic healthcare record data to investigate several aspects of the epidemiology of diabetes in Scotland. My most recent work in this area has investigated rates and predictors of acute complications of type 1 diabetes.

I am also employed on the IMI funded hypo-RESOLVE project as a data analyst. My role in this project is to design and perform statistical analyses using clinical trial data that has been pooled from multiple providers in the pharmaceutical industry. For this project I am investigating the quantitative association of predictors of hypoglycaemia with a range of hypoglycaemic outcomes, and quantifying the clinical effects of exposure to hypoglycaemic events.


I obtained a PhD in Bayesian phylogenetic methods from The School of Earth Sciences at The University of Bristol in 2017. I was subsequently employed as a postdoctoral researcher in the group from 2017 to 2019. My research interests in phylogenetics are model development and assessment of model efficacy. I published several papers investigating aspects of Bayesian divergence time estimation, Bayesian/frequentist and non-parametric topology estimation, and the empirical applications of such methods.

While my day to day research is no longer focused on phylogenetic methods, I still retain a keen interest in developments within the field.

Key Skills

Programming And Computing Tools

I use some form of programming everyday as part of my work. I have ~9 years of experience with R and it is the language I use most frequently. I have an strong understanding of base R and it’s more advanced features and I am also proficient with a range of commonly used R packages. For most data manipulation tasks I tend to prefer data.table over tidyverse. See below for a non-exhaustive selection of hexstickers for R packages I regularly use:

I have several years experience of MySQL, postgreSQL, and Oracle SQL, for querying the relational databases I work with. I also have ~ 10 years of experience using Unix and Linux environments and I am well versed in remote access approaches such as ssh, and the use of high performance computing (HPC) environments. I have some experience in python, C++, JavaScript, html, and css. I can also write code in the probabilistic programming language stan. For version control I have used both git and subversion. My github profile can be found here, due to governance rules surrounding the data I use for my job I only commit non-work related repositories to github.


In my research I have applied a range of methods and have become proficient in their use. In my day-to-day research I use regression based modelling for both causal inference and predictive modelling. I am experienced in the use of generalised linear models, generalised linear mixed models, and generalised additive models. I am also experienced in the use of Bayesian inference, particularly in the contexts of hierarchical regression modelling and phylogenetic methods. I also regularly use time-to-event models and apply joint modelling of longitudinal and time-to-event models.


I have authored over 20 peer-reviewed articles and have given oral presentations of my work at dozens of conference meetings all across the world. I have presented my work at SVP, ICP, GSA, EASD, Diabetes UK, The Royal Society, PalAss, and GRC. For the 2017-18 academic year I organised and hosted the Bristol Palaeobiology Research Group’s seminar series.

For data visualisation, I am proficient with ggplot2 when using R. I can also construct web-based visualisations in Vega-Lite and I have some experience with visualising data with d3.js.


I was a guest lecturer for Palaeobiology course (MSc/MSci) at The University of Bristol (2017-18; 2018-19) for modules “Phylogenetics” and “Current Controversies in Palaeobiology”. I was also a personal tutor for MSc students (2017-19). In 2018 I was part of a delegation from The University of Bristol that travelled to NIGPAS - Nanjing, China to run a workshop “Workshop on Molecular Palaeobiology: Phylogeny and Divergence Time Estimation” on phylogenetic methods. As part of this delegation I ran practical workshops and delivered lectures.


My CV is available as a PDF here.