CABM banner

Leverhulme Centre for Advanced Biological Modelling

The Leverhulme Centre for Advanced Biological Modelling (CABM) will harness cutting edge mathematical and computing skills to address major problems in biology. Biology at the University of Sheffield is defined by its strength in combining approaches across time-scales from ecological to evolutionary, and has been the subject of significant recent investment. Exploiting this breadth of research excellence, the areas of focus for the CABM will extend our cutting edge research in basic biology by harnessing the latest mathematical and computational techniques.

The recruitment process of the CABM will focus on attracting outstanding students from physical and mathematical sciences, and training them to address leading biological problems. This multi-disciplinary initiative will transform our research excellence in organismal biology and fill a chronic training gap in biology. Its legacy will be a cohort of research scientists at the cutting edge of biological modelling with the skills to tackle major societal problems.

These projects are available to UK, EU and overseas applicants, but funding for tuition fees will be provided to successful applicants at the UK/EU rate. This means that students from outside the EU will need another source of funding to cover the difference between UK/EU tuition fees and international fees.

Funding your PhD

To apply for any of these projects, you should complete the University of Sheffield's postgraduate online application form.

Postgraduate online application form

Modelling genomic and phenotypic evolution in a natural population






We will not achieve an adequate understanding of adaptation in natural populations until we can model the interaction among the genome, the environment and individual organisms, and so predict the phenotypes and genomic composition of subsequent generations. This project will exploit the Flexible Large Scale Agent Modelling Environment for the Graphics Processing Unit (FLAMEGPU) agent-based modelling (ABM) environment, developed by Paul Richmond, to model the evolution of individual reproductive success in the Lundy population of house sparrows. FLAMEGPU is open source and provides an abstraction level that allows modellers to focus on specifying agent behaviour, whilst providing automated code targeted to GPU architectures that provide accelerated levels of computational performance. The Lundy Sparrow Project, led by Terry Burke, has collected detailed phenotypes and complete individual life-histories of individual sparrows for more than 16 years, comprising many generations. The study will combine an open-source genetic modelling package that allows recombination (based on, e.g., FIGG or GPOPSIM) with the ABM to simulate the evolution of whole genomes in the population through time. The model will be tested by comparing the genomes and phenotypes of the current population with those in the starting population. The project will be unique in combining and ABM with whole-genome data, and in combining population dynamics, population genetics and real-world fitness data.


Terry Burke (APS) will provide the data to be used in the project and expertise in population genomics and avian life history.

Paul Richmond (Computer Science) will provide expertise in state of the art agent-based modelling and computer programming, especially using Graphics Processing Units (GPUs).

Mark Rees (APS) will provide expertise in demographic and life-history modelling, including the statistical verification of model performance.

Predicting Genomic Effects on Structural Colour

Structural colours, caused by diffraction or interference of light from nano-scale structures, are some of the brightest and most impressive colours in nature. There has been extensive interest in replicating structural colours, which are produced from simple biological materials that are cheap and non-toxic, and can also have unique properties such as angle dependent colour (iridescence). However, very little is known about how the patterning of such fine structures is controlled in biological systems and the genetic pathways controlling this have yet to be identified in any system. We have performed crosses between subspecies of Heliconius butterflies that differ in iridescent structural colour and are using these to identify genetic loci underlying these differences. We also have genomic data from natural hybrid zones between these subspecies, which can be used to identify narrower genomic regions. This project would use this genomic data to predict how individual genetic loci control specific aspects of structural variation, and so fit within the area of “post-omic forecasting”. The project will involve computational modelling approaches, firstly to predict fine structural characteristics from small angle x-ray scattering data and secondly to predict this morphology from genotype.


Nicola Nadeau has extensive experience of analysing genomic data to characterise patterns of divergence and gene-flow between species, as well as to identify genetic variants underlying phenotypic variation through approaches such as genome-wide association. She has authored several high profile publications identifying genetic variants underlying variation in colour and pattern in both butterflies and birds. This project would build on her NERC fellowship, using traditional genetic mapping approaches to identify the genetic architecture and loci underlying the convergent evolution of iridescent structural colour in two species of Heliconius butterfly. The biological aspect of this project would involve using existing genomic data from her lab, to identify genetic loci controlling specific aspects of fine structural variation.

Andrew Parnell is a physicist with a background in structural characterisation, using real space (scanning probe microscopy, optical) and reciprocal space scattering techniques (neutron and x-ray scattering). Throughout his career he has been interested in the dynamics of structure development in soft matter polymer and biological systems, with published and continuing studies to understand the nano-structuring and development of the structural colour elements of Jay feathers barbs and ultra white beetle scales. The physics element of this new project will involve using a combination of different precision structural techniques to correlate the differences in iridescent structural morphology with the underlying genetics responsible for their development. This will use scanning probe microscopy to directly image the local surface of the iridescent butterfly scales, alongside scanning electron microscopy for larger areas. These techniques give an overall picture of the diversity and variance in surface structure. We will then use ultra small angle x-ray scattering (USAXS), available to us at Sheffield via the recently installed Xenocs liquid gallium beamline, which will allow us to access the length scales relevant to structural colour in three dimensions. We also plan to use the high throughput facilities available to us via our successful and ongoing USAXS programme at the instrument ID02 (European Synchrotron Facility (ESRF)) to obtain structural data for all genotyped individuals.

A Hyper-Omics approach to reveal how diet reduces ageing

Living longer and, especially, healthier is a key objective of medical science and a growing priority in our ageing society. The only known reliable option to achieve this is dietary restriction. When the results from model organisms are extrapolated to humans, interventions or drugs that mimic dietary restriction are predicted to lead to life- and health-span extension that is larger than curing all cancer and cardiovascular disease. Hence, the impact of understanding the mechanisms underlying dietary restriction for medical science is immense, yet progress to uncover these mechanisms has been limited. In addition, on a fundamental theoretical level, the mechanisms of ageing underlie all major life-history trade-offs and understanding ageing is thus key to understand many areas of biology.
We propose a novel Hyper-Omics approach in flies (Drosophila melanogaster) to reveal the mechanisms of dietary restriction. Flies are a convenient model, because dietary restriction reduces mortality risk strongly (over 5-fold) and instantly (within a day). Furthermore, flies have a comprehensively annotated transcriptome, proteome and genome available; and candidate genes can be tested readily using transgenic flies.
Datasets will be generated that consist of full (phospho-)proteomes, metabolomes and transcriptomes in a time-series after exposure to dietary restriction. Combined inference from different omics is in its infancy, but has great promise to understand complex biology. You will develop the first complete framework for such combined inference and will extend this using molecular evolution approaches. Statistical techniques can involve (amongst others) principle-component-analysis, meta-analysis, molecular evolution and Bayesian Network analysis. Experimental results from tests of candidate genes identified (Simons’ lab) will further inform the developed framework to provide a detailed insight into the complex biology of dietary restriction. You will require a strong quantitative skillset, but direct experience in any of these statistical techniques or in biology is no requirement to be considered for the studentship.


Mirre Simons (lead) is a Vice-Chancellor’s fellow and works from a combined evolutionary and biomedical perspective on the biology of ageing. He uses comparative, theoretical and meta-analytic methods to study the process of ageing and uses Drosophila melanogaster (the fruit fly) as a powerful genetic model system. Mirre has a strong statistical skillset with a proven track record of developing novel statistical approaches to understand the biology of ageing. Mirre’s lab will provide the data required for this project and can directly test candidates revealed by the efforts of the student. This ensures a direct embedding of the work of the student in actual biology and empirical feedback through experimental testing. Mirre will further specifically support the statistical approaches to develop the combined inference methods.

Toni Gossmann (co-supervisor) is a Leverhulme Early Career Fellow with a background in bioinformatics. He works on theoretical aspects of evolutionary biology using the latest techniques in proteomics, genomics and transcriptomics. He has experience in the application of several computational approaches in sequence analyses using molecular evolution and population genetic frameworks. His key interest is to understand how evolution operates on the genomic level, but also uses novel tools to understand complex biology. A key aspect of this is reflected by his work on NAD metabolism, a fundamental component in ageing, and an important field in metabolomics. Toni will advise on the statistical approaches, on the generation of sequence datasets and will specifically support the student regarding the computational aspects of the project, in particular the molecular evolutionary approaches that are crucial for the understanding of the genetics of the underlying traits

Interrogating the great tit (Parus major) genome for evidence of gene flow in a hybrid zone in the Far East

Being able to accurately quantify the pattern of gene flow between populations is essential for understanding the process of population divergence, local adaptation, and ultimately speciation. As part of the ongoing population/ecological genomic studies of great tits in our labs, this PhD project intends to investigate the genomes of great tit samples collected from an area in the Far East where they coexist with a closely related species, the Japanese tit (Parus minor), and apparent hybrids can be found. However, very little is known about the evolutionary relationships between the European great tits, the hybrid zone population, and the Japanese tits, and it is unclear what role gene flow has played in shaping their evolution. To address these questions, the student will assist in testing the new methods for detecting gene flow that are being developed in our labs, and will apply them, and other well-established inference methods, to whole-genome DNA sequence data collected from these populations. The project is multi-disciplinary, combining methodological developments and analysis of real data. It is suitable for applicants who have backgrounds and training in physics, maths or computer science or those with advanced numerical skills from a biology background (e.g. through masters-level training).


Kai Zeng (KZ) is a population geneticist with a strong interest in using large-scale DNA sequence data to understand how fundamental forces such as natural selection and demographic changes drive evolution. With a background in statistics and computer science, a significant part of KZ’s research has been the development of models and statistical inference procedures for analysing data. More related to this project, KZ’s lab was heavily involved in the Great Tit Genome Project, and has been conducting population genomic analyses using whole-genome data acquired from great tit samples collected from Europe. KZ will be mainly responsible for leading the methodological part of this PhD project.

Jon Slate (JS) is an evolutionary geneticist with a particular interest in using molecular tools to understand the genetic architecture and microevolution of Mendelian and quantitative traits in natural vertebrate populations. JS has been working on great tits for many years, was a key contributor to the Great Tit Genome Project, and has been leading the Great Tit HapMap Project. In this PhD project, JS will be responsible for providing biological insights into to the modelling and the interpretation of results.

Automated linking of genotype and environmental covariates to predict the effect of landscape change on evolution

The recent explosion in wild populations’ genomic data enables vast improvements in our understanding of how genotypes affect a wide range of biological phenomena. Research has successfully linked genotypes to longitude and latitude coordinates (Elhaik et al. 2014, Nature Communications), but to understand the role of phenotypic plasticity in adaptation to different geographical situations, it is important to link genotypes with other environmental-geographical factors, such as altitude, habitat type, and climate.
We will build on our models (e.g., Elhaik et al. 2014a; Das et al. 2016; Marshall et al. 2016): by 1) Widening the genotype-environment link to incorporate a variety of published geophysical, ecological and meteorological data, 2) Making them species independent, 3) extending them to haplotype-level. The resulting gene-landscape correlations will enable us to predict which individuals are more likely to survive under landscape- and climate-change scenarios, therefore determine the likely evolutionary path a population would take.
This project will exploit existing two major genomic/environmental/phenotypic datasets for humans and Medicago truncatula. It will run alongside Elhaik’s recently funded NSF/MRC/Industry grants for developing tools to infer human ancestry from genomics. The student will collaborate with anthropologists at Central Washington University and develop highly demanded cutting-edge bioinformatics tools


Eran Elhaik is a computational biologists in The Bioinformatics Hub and The Department of Animal and Plant Sciences at the University of Sheffield, whose research focusses on complex traits (or disorders) and population genetics. As the primary PI, his role will be in training the student in genetics and introduce relevant bioinformatics tools and methods necessary to carry out the tasks. He has much experience analyzing non-human (Elhaik, Pellegrini, and Tatarinova 2014) and human (Elhaik et al. 2014a; Elhaik et al. 2014b) population genetic data and studying complex disorders (Elhaik and Zandi 2015).

Jonathan Potts is a mathematical ecologist based in the School of Mathematics and Statistics at the University of Sheffield, whose research focusses on spatial ecology. His role will be in advising the student how to build and parametrise mathematical models of the spread of genes through heterogeneous, changing environments. This will involve adapting tools he has developed (Potts et al. 2014a,b, in press) for modelling the smaller time-scale problem of animal movement though heterogeneous landscapes.