Optimal Real-Time Decision-making

in collaboration with National Grid

With the changing generation mix, demand behaviour, and technology landscape in the electricity industry, better planning tools to study and control the behaviour of the system are becoming more important. One of the areas of interest is to be able to use the output of monitoring tools such as Phasor Measurement Units to make operational decisions in real time. In the future, decision-making based on the output of a number of these devices on the system to enhance system performance is desirable. The purpose of this research project is to develop mathematical models which facilitate developing new control tools based on multiple-input control functions.

LTE femtocells for future wireless networks

in collaboration with BT

The wireless research team at BT Technology is increasingly interested in LTE femtocells. These are small 4G base stations, which would be used by domestic users as a hub for all networking services – voice, internet, and television. These femtocells already exist and work well in isolation, but the deployment of large numbers poses difficult problems in the area of distributed management of network resources, and some of these problems are highly mathematical. The difficulties largely come from the fact that centralized management is impossible, and we need a self-organizing network (SON). Key questions are (i) Interference management (ii) Self adjustment of devices (iii) Time synchronization (iv) Multi-input Multi-output using multiple radio paths.

Response to Epidemics

The Bioterrorism and Emerging Disease Analysis programme provides evidence based analysis and advice to Public Health England, Dept. of Health or other bodies to inform both preparedness for, and the response to, the public health threats arising from bioterrorism and new or (re-)emerging infectious diseases. Typical problems might arise from natural releases of biological contaminants, response to bio-terrorism, habitat and climate change or planning for, potentially sparse, pharmaceutical resource allocation. Research questions range from the micro-scale (How do we model the dose response to a biological agent and effect of countermeasures) through to macro-scale (How do we mitigate a epidemic impacting the national scale) and consider behavioural drivers as well as physical logistics of a response.

Integration and processing of noisy signals for decision making

We have recently developed a theoretical framework for the way in which cells transmit information from the external environment to their genes and then use this for decision-making, for example to decide to divide, die or change their differentiation state. This involves a combination of ideas from dynamical systems, information theory and hypothesis testing. In this project we will develop this further so as to generalise the range of application to technological application such as the processing of noisy signals in cars. The theory will be developed alongside its application to one of these technological areas.

From neurons to neuronal fields

The complex patterns of cortical activity – up-down states or epilepsy - are the focus of a massive international research effort. Recently there have been significant advances in our knowledge of the cellular components of cortical tissue, the dynamical properties of the synapses that connect them, and an increasing amount of data available at the mesoscopic level, at which the collective activity of many thousands of neurons can be measured. This project proposes to use mathematical and computational modeling to understand how the biophysical properties of cells and synapses lead to the emergent states seen at the level of networks. This theoretical project will benefit from close collaboration with the experimental group of Dr Mark Wall from which electrophysiology and wide-field calcium imaging data will be used to constrain the models, with the model predictions feeding back into the design of future experiments.

Evolution in households

Competition between strains of infection is key to the evolutionary dynamics of infections, which in turn determines how infections response to changes in their environment, such as recent changes in human population structure. The vast majority of evolutionary models assume simple mean-field (or random-mixing) for the transmission dynamics. In reality transmission is shaped by the local social environment, with households playing a major role in shaping and constraining transmission. This close clustering of contacts increases local competition for available susceptible individuals which in turn may shape evolutionary pressures. Recent methodological advantages allow us to rapidly model the spread of infection within households and these models could be extend these to include multiple cross-reactive strains. The aim is to understand evolutionary pressures acting on human pathogens and how these may have been modified by recent changes in household structure.

Spatial heterogeneity in networks of smooth muscle cells

The uterine wall muscle is composed of a network of smooth muscle cells. An intriguing property of these cells is that their repertoire of ion channels is highly redundant, meaning that a cell is capable of realizing the same biophysical behaviour using different "minimal" subsets of the entire set of ion channels that is evident from the tissue-level transcriptome. It is at present not clear whether these cells do in fact choose from among the minimal subsets, or whether each cell expresses the entire repertoire. The former possibility might confer an enhanced ability to modulate excitability at the whole-organ level; it is of great interest to investigate this since any avenue to fine-control uterine excitability is a potential means to treat conditions such as pre-term delivery, pre-eclampsia, and post-partum hemorrhage.

Patient stratification from phenotype and genotype data

in collaboration with CRUK and University Hospitals Birmingham

The CDT's partnership with the University Hospital Birmingham, the CRUK Centre in Birmingham and oncologists in Birmingham and Warwick opens up the opportunity for a project combining molecular level systems biology and healthcare sector data such as electronic patient records and biobank questionnaires. The aim is to group and stratify patients not only from their genotype, but also phenotypically based on the clinical descriptions in the medical records. Combining molecular level and healthcare sector data and providing the models and tools to analyse this is challenging but progress in this largely unexplored area is likely to have a great impact.

Differential Footprinting of DNase-seq data

in collaboration with Birmingham School of Cancer Sciences

DNase-seq is a state-of-the-art approach to analyse the global pattern of protein-DNA binding in regulatory regions of DNA. In this assay, DNA is digested by DNaseI and high-throughput sequencing is applied to reveal the genome-wide pattern of DNA digestion in hypersensitive regions. Regions of DNA bound by regulatory proteins are protected from digestion and leave behind 'footprints', regions of reduced cut frequencies. Paired DNase-seq data sets are challenging to analyse, but provide means to perform a comparative analysis of regulatory events. This becomes important, when, for example, the regulation of stem cells and differentiated cells or normal and diseased cells is being compared.

Swarm computation and swarm-like materials

Swarming is familiar in many animal systems, e.g. bird flocks, fish shoals, insect swarms, human crowds etc. It is an important model for social interactions and has applications in computing, robotics, pest control, animal conservation, crowd control and panic management. Swarming also occurs in thermodynamic systems in which particle motion responds to gradients in fields such as temperature or concentration of a solute species. These fields can themselves be controlled by the distribution of particles, resulting in a rich space of dynamical solutions that we are only just beginning to understand. One urgent priority is the development of continuum models of systems with long ranged interactions. A second is an extension of models of animal behaviour to include the effects of very short term memory. Finally swarm-like materials also exist in systems with physically connected components connected by motors that drive them out of equilibrium. These new materials represent novel physical systems and may have numerous industrial applications.

Statistics of the Microbiome

in collaboration with School of Life Sciences

The traditional view of one infection - one host has been over-turned by the wealth of genetic data collected on the microbes found on and within our bodies. Recent advances have shown that multiple infections are the norm and made actually be necessary for healthy living. Disease may therefore be viewed as an imbalance in this collection of microbes. Work in the School of Life Sciences at work has focused on mastitis, and the microbiome found with the udder. Novel statistical methods need to be developed to deal with the large and complex amounts of data generated by such projects to provide insight into the causes of disease.

Dynamics and persistence of seasonally forced infections

It is well known that many infections show pronounced multi-annual cycles due to the presense of seasonal-forcing in the transmission rate. For example, measles (before vaccination) showed a clear 2-year pattern that can be captured by age-structured models with seasonal forcing due to school terms. Such models could be readily extend to other disease parameters to quantify how the periodicity changes. In addition, stochastic version of such models could be used to examine persistence of infection under a variety of conditions such as (i) how does this pattern of persistence / extinction change for a range of disease parameters? (ii) what is the impact of adding extra structure, such as spatial structure, schools or homes? Understanding the persistence and re-emergence of infection are key tools in disease eradication.

Systemic Risk in the Financial System

in collaboration with Sciteb

The European Banking Authority published the results of a stress test of 123 of the EU's banks on 26 October 2014. The project is to deduce the implications for resilience of the banking system as a whole. As was acknowledged by leading UK economists after the 2008 crisis, "the failure to foresee the timing, extent and severity of the crisis and to head it off, while it had many causes, was principally a failure of the collective imagination of many bright people, both in this country and internationally, to understand the risks to the system as a whole." Similarly, Andy Haldane, director of financial stability at the Bank of England said "[A] conceptual problem with [Basel Accords] risk-weighting is that it takes no account of the collective consequences of banks' asset allocation decisions."