UCL CoMPLEX

William Ashworth ^

Designing a bioartificial liver device

Supervisors:
Dr Nathan Davies (Hepatology)
Prof David Bogle (Chemical Engineering)

Liver disease is the fifth most common cause of death in the UK, and the only major cause of death rising year on year. Liver transplants are performed at a rate of about 700/year in the UK and last year 100 people died while waiting on the list. With 20,000 new diagnoses of cirrhosis and 50,000 hospital admissions due to paracetamol poisoning every year, a functional bioartificial liver (BAL) device for acute and chronic liver patients as a bridge to recovery or transplantation remains and imperative but unmet clinical need. Current devices use flat plate hollow fibre membrane and cell scaffold designs incorporating hepatocyte cell lines but none have shown significant clinical efficacy. The key problem with current systems is the absence of cell-cell signalling for normal hepatocyte function.

The research goal is to create a BAL system using a cryogel, a polymeric irregular porous 3D support. The cryogel is potentially advantageous since it allows a much greater surface area for cell/blood contact and mimics the natural structure in the liver. It allows adhering cells to grow and perform metabolic and signalling functions for more realistic hepatocyte function and allows blood processing without pre-treatment. Computational modelling will be required to ensure the system permits sufficient cell loading with adequate mass transfer of oxygen, carbon dioxide and nutrients for prolonged hepatocyte replacement of liver function. Furthermore, since hepatocytes removed from the liver are fairly inefficient and have a high rate of mortality, modelling will be used to optimise cell-cell signalling, maximising the survival time of the hepatocytes and their ability to perform normal liver function in the BAL device.

www.ucl.ac.uk/~zcapd06

Chris Banerji ^

Network Theoretic Tools in the Analysis of Complex Diseases

Supervisors:
Dr. Simone Severini (UCL Computer Science)
Dr. Andrew Teschendorff (UCL Cancer Institute)

Collaborators:
Prof. Peter Zammit (KCL Randall Division of Cellular and Molecular Biophysics)
Dr.Richard Orrell (Royal Free Hospital, Neurology)

We are interested in elucidating the functional pathomechanisms of complex diseases, for which either the genetic basis is poorly characterised and/or for which the genotype to phenotype link is highly non-trivial. We are developing and applying novel network theoretic techniques, many of which are rooted in information theory for achieving this aim. We are currently focused on breast cancer and fascioscapulohumeral muscular dystrophy (FSHD). For the former we are interested improving the understanding of the role of nuclear receptors, particularly in identifying nuclear receptors driving tamoxifen resistance in ER+ breast cancer, and important nuclear receptors which may prove drug targets in triple negative breast cancer. The 47 human nuclear receptors modulate the expression of a large number of downstream targets in a highly combinatorial manner, orchestrated via an intricate dimerisation network, we believe that the application of carefully developed network theoretic tools may shed light on potential drug targets that would otherwise be over looked.

We are particularly interested revealing in more detail the pathomechanisms of FSHD, a poorly characterised, yet the most prevalent form of inheritable muscular dystrophy, for which there currently exists no viable treatment. This genetic basis of FSHD is well characterised, and the primary mechanism is widely accepted as the aberrant expression of a novel transcription factor (DUX4).
However, how DUX4 manipulates genetic programmes to drive the highly varied FSHD phenotype is poorly understood. Through a collaboration with the Zammit lab and Dr. Richard Orrell, we will couple state of the art molecular biology with powerful mathematical tools to identify druggable pathways for this highly complex condition.

www.ucl.ac.uk/~ucbpban

Robert Bentham ^

Bioinformatic analysis of the interface between mitochondrial biogenesis and cell death signalling pathways

Supervisors:
Dr. Kevin Bryson (Computer Science)
Dr. Gyorgy Szabadkai (Cell and Developmental Biology)

Mitochondria evolved to endow eukaryotic cells with adaptive potential by calibrating cellular energy levels to specific needs both under normal and stress conditions. Moreover, the organelle is a hub of cellular signalling pathways regulating cell death and survival. Thus mitochondrial dysfunction lies at the heart of several major pathologies such as neurodegenerative and chronic cardiac disease, diabetes and cancer. The maintenance of a competent cellular mitochondrial population is achieved by the balance between (i) selective degradation of damaged organelles by autophagy and their (ii) replacement by mitochondrial biogenesis, i.e. transcriptional regulation of nuclear encoded mitochondrial proteins. While recently the role of mitochondrial autophagy in disease has been extensively studied, little is known about mitochondrial biogenesis in this respect.

Big data sets such as the Cancer Cell Line Encyclopedia and other large microarray datasets will be used to examine gene expression under different conditions and disease states. Various sophisticated computational techniques ranging from machine learning methods (Sparse linear models, random forest etc.) to bayesian statistics will be applied to these datasets in order to elucidate the relationship between mitochondria biogenesis and cell death signalling pathways. By investigating the transcriptional regulation of apoptotic and necrotic cell death pathways by the PGC-1 familiy of coactivators, the final aim is to identify a novel pathway linking mitochondrial biogenesis and cell fate.

www.ucl.ac.uk/~ucbprbe/

Katharine Best ^

Analysis of the T-cell receptor repertoire

Supervisors:
Prof. Benny Chain
Prof. John Shawe-Taylor

The adaptive immune system generates an enormous diversity of receptors on T and B lymphocytes, each with the potential to stimulate a specific immune response to a particular pathogen.  The size of receptor diversity is estimated to be in the order of 109 - 1010 different receptors, and massively parallel high-throughput sequencing has allowed direct analysis of this repertoire. This project will involve a mixture of molecular biology laboratory based experiments, focussed on developing improved protocols for TcR sequencing using an in cell PCR protocol, and computational work. I will attempt to development novel computational tools to validate, store and analyse the complex datasets generated, with the aim of providing quantitative data on clone size, clone diversity and kinetics during immune responses.

Rollo Hoare ^

Immune reconstitution in children

Supervisors:
Robin Callard (Institute of child Health)
Joseph Standing (Institute of Child Health)

I will be looking at the reconstitution of the immune system in children after a bone marrow transplant (BMT) and under antiretroviral therapy (ART) with HIV. Patients undergo a bone marrow transplant (BMT) as treatment for cancer or immune system failure. Before a BMT, patients may undergo conditioning that causes the complete ablation of their immune system. I will be using non-linear mixed effects (NLME) modelling, at first to look at which covariates affect the reconstitution using a simple empirical model. Later I hope to develop a more mechanistic model that will more closely resemble the underlying system, allowing greater understanding of the reconstitution and of how the reconstitution is affected by the covariates.

Similarly, in children with HIV, the concentration of CD4+ lymphocytes is significantly lowered by the HIV virus, and under ART, this concentration can increase significantly to an improved homeostatic level. We can use NLME modelling to understand what affects this recovery. I will be assisting in the development of mechanistic models to aid our understanding of the reconstitution.

www.ucl.ac.uk/~ucbprlh

Nargess Khalilgharibi ^

Investigating the short timescale mechanics of living monolayer tissues

Supervisors:
Prof. Mark Miodownik (Mechanical Engineering) Dr. Guillaume Charras (LCN: London Centre for Nanotechnology)

Multicellular organisms are made of different types of tissues, the simplest of which are one-cell-thick monolayers. Monolayer tissues play an important role through the lifetime of the organism, i.e. during the early development of the embryo and later in the physiology of the organism. Although their function involves generating internal forces and resisting to external forces, a deep understanding of their mechanical properties is lacking.

When put under stress, one-cell-thick monolayers change shape and elongate, while becoming thinner. Force relaxation experiments have revealed that under high strain rates, the response of tissue monolayers to mechanical stress is fast, leading to a significant change in shape (e.g. elongation). In my PhD, I aim to combine experimental and computational techniques to explore the origin of this mechanical response and discover how much of it is purely passive. The experimental part of my project incorporates implementation of a high sampling rate measurement technique in the existent system using force transducers. This will be accompanied by developing a FEM model, which will provide the theoretical framework for understanding the experimental results.

www.ucl.ac.uk/~ucbpnkh

Tim Lucas ^

The macroepidemiology of Nipah and Ebola viruses in bat populations in Ghana

Supervisors:
Kate Jones (Genetics, Evolution and Environment/Institute of Zoology) Hilde Wilkinson-Herbots (Dept. of Statistics)

Zoonotic diseases make up a large proportion of emerging diseases globally. Bat populations in Africa have been shown to carry a number of important diseases including Ebola and Nipah virus. While spillover of Ebola to humans has occurred a number of times in Africa, Nipah virus has so far only infected humans in SE Asia and Bangladesh. The ability to predict if, when and where these viruses will spillover to human or livestock populations will allow for effective monitoring and efficient responses to outbreaks.

As these diseases are intrinsically tied to their fruit bat hosts, any effective model must start by understanding the ecology of these species. This project will use landscape-scale ecological modelling, epidemiological models and data collected from the field to elucidate where these diseases are most likely to infect human populations.

www.ucl.ac.uk/~ucbptcl

www.twitter.com/timcdlucas

Daniel Manson ^

Electrophysiological studies of grid cell behavior in freely moving rats

Supervisors:
Prof. John O’Keefe (Cell and Developmental Biology) Prof. Neil Burgess (Institute of Cognitive Neuroscience)

Using metal electrodes (with a diameter of ~20 microns) it is possible to record the activity of neurons in a chosen region of the rat brain.  I am interested in the brain's representation of spatial information, so I do these recordings while the rat is running around a small arena looking for rice.  In the hippocampus and entorhinal cortex there are several types of neuron known to encode spatial information, namely head direction cells, place cells, and grid cells.  Head direction cells seem to be tasked with maintaining an internal compass; place cells seem to act as internal labels for specific locations within the arena; and grid cells seem to label the points in a rhombus that is then tessellated across the arena forming a repeating grid.  At the start of my PhD I will be examining how the scale of the grid changes as the rat familiarizes itself with a new arena.  Later on I hope to examine whether grid cells deal with spatial information in general or only with self-location.

www.ucl.ac.uk/~ucbpdma

Elizabeth Moorcroft ^

Assessing the efficacy of camera-trap survey methods using models which incorporate biological factors

Supervisors:
Prof. Stephen Hailes (Computer Science)
Dr. Marcus Rowcliffe (Institute of Zoology)
Dr. Chris Carbone (Institute of Zoology)

Estimating the number of animals remaining within a population and their spatial distribution is fundamental for both conservation and management of endangered species. Many of the assumptions used in the current methodology are rarely met in real populations, and are particularly unrealistic when species are under considerable stress - such as in large felines.

This project aims to explore the relationship between animal movement, behaviour and population estimation. The project will develop new techniques for assessing available tools and new methods for incorporating our understanding of behaviour, movement and environmental variation into population estimation. Validation of these methods will be conducted through existing or collaborative field study via the Zoological Society of London.

Kim Mroz ^

Understanding and Characterising Biodiversity Sound with Machine Learning

Supervisors: Prof. Kate Jones (GEE) Prof. Mark Girolami (Statistical Sciences) Gathering visual data on various animal populations can be extremely difficult, not to mention regulated. An alternative is to collect audio recordings, which for particular taxa can yield a wealth of information. For instance, many animals may emit species-specific calls and if that animal is also small or fast or rarely seen then reliable acoustic identifiers would greatly assist in biodiversity and ecology research. Existing studies have resulted in vast amounts of audio data and various machine learning techniques have been applied, producing acoustic classifiers for various terrestrial and marine groups such as birds, insects, bats and whales. This project aims to explore and develop learning techniques to improve upon the reliability of current classifiers, while also exploring the possibility of a general biodiversity classifier, unifying the methods across taxa.

www.ucl.ac.uk/~ucbpkmr

Shailendra Rathore ^

The Development of the Spatial Representation System

Supervisors:
Dr Francesca Cacucci (Institute of Behavioural Neuroscience)
Prof. Neil Burgess (Institute of Cognitive Neuroscience)

The mammalian brain represents the spatial location and orientation of ananimal relative to its environment in the firing of place, head-direction, grid and boundary-vector cells.
We have computational models for these systems, yet how such patterns of activity can be learnt is an enduring question. The aim of my project is to study the development of the spatial-navigation system through a synthesis of studying unsupervised learning rules in neural models and using in-vivo electrophysiology in rat-pups as they navigate out of the nest for the first time. The development of the spatial firing of each of these cells could provide fundamental details of the formation of more downstream firing patterns and illuminate roles of these cells in encoding spatial trajectories.

Víctor Sojo ^

Bioenergetics at the origin of life

Supervisors:
Dr. Nick Lane (GEE: Genetics, Evolution and Environment)
Prof. Andrew Pomiankowski (GEE)

Although a universal definition still escapes us, life can be described as a far-from-equilibrium process that requires compartmentalisation, metabolism, heredity, and the input of energy and materials coupled with the ability to excrete waste.

The Origin of Life (OoL)  is a matter of chemistry, specifically the transition from geochemistry to biochemistry. Hypotheses abound, from Darwin’s "warm little pond" to deliberate inoculation from outer space. However, most scenarios overlook key aspects of the chemistry of a system that, as all life, must have been in continuous thermodynamic disequilibrium.

The last two decades have seen the development of a theory that fulfils all conceptual requirements, but an understanding of the exact chemistry is lacking. Mike Russell and others have introduced what seems to be the first fully congruent scenario for the origin of life in a purely inorganic geological system, namely a specific type of alkaline submarine hydrothermal vent. The proposed vents are formed by the process of serpentinization of ferrous-magnesium minerals like olivine in the oceanic crust, which produces hydrogen-rich alkaline fluids that permeate towards the carbonated acidic waters of the ocean above. In the oxygen-devoid waters of early Earth, solutes precipitated forming cell-sized micro-compartments rich in minerals like the Fe-Ni-S-containing mackinawite, conceivably the first catalysts for life.

I plan to use a combination of phylogenetics and analysis of modern biochemistry to study the origins of key metabolic pathways in the dawn of life and up to the Last Universal Common Ancestor of all beings alive today, fondly known within the OoL community as LUCA.

www.ucl.ac.uk/~ucbpvso

Claire Walsh ^

Bubble behaviour in human tissue and its consequences on
decompression sickness

Supervisors:
Umber Cheema (Division of Surgery & Interventional Science)
Nick Ovenden (Mathematics)
Dr. Eleanor Stride (Institute of Biomedical Engineering)

Decompression sickness (DCS) commonly known as ‘the bends’ is a condition
predominantly suffered by deep-sea divers. It is caused by bubble formation in tissues throughout the body and can have range of consequences including joint pain, seizures and death.

Current DCS prevention schedules are based on underlying theory that is over 100 years old, and scant consideration has been paid to the physics of bubble formation or diver physiology.

An industry driven need for better prediction of safe dive schedules has lead to a partnership between VR Technology and UCL to attempt to answer some of the more fundamental questions regarding where and why bubbles form in tissue.

A model of multiple neighbouring gas bubbles under decompression has been
developed by Jean-Pierre O’Brien at UCL in partnership with VR Technology. My PhD aims initially to validate this model using collagen type I tissue scaffolds; and then further develop it to include tissue parameters such as increased vasculature, levels of hydration and effect of mechanical stress and strains, which can be controlled within these scaffolds.

Thomas Wyatt ^

Stretch and grow: Regulation of cytokinesis in epithelial tissues by mechanical forces

Supervisors:
Dr. Guillaume Charras (London Centre for Nanotechnology)
Prof. Buzz Baum (Laboratory for Molecular Cell Biology)

Cells are able to monitor the mechanics of their surroundings and alter their own mechanical properties in numerous ways. In recent years it has become clear that a cell's ability to detect and respond to forces plays a fundamental role in many important biological processes such as tissue morphogenesis and cell differentiation. In this project I will be adapting newly developed tools which apply mechanical stress to monolayer tissues in order to study the effect on cell division. Previous work in the laboratory has shown that mechanical stress can trigger an increased rate of division and that divisions are oriented along the axis of stress. Ultimately we would like to understand the molecular mechanism behind this process and understand its role in tissue development and homeostasis.

www.ucl.ac.uk/~ucbptpj/