Seedcorn projects

PIs: Niki Karavitaki and Gabriela da Silva Xavier. Co-I: Alexander Zhigalov
Artist in Residence: Lucy Hutchinson

Acromegaly is a serious medical condition caused by a tumour in the brain secreting by too much growth hormone and leading to increased growth and damage to the body. As a result of acromegaly, patients also have a high risk of diabetes, heart attack and premature death. Getting treatment right is very important but the success of treatment is difficult to measure, meaning it is difficult to know if treatment is effective. 

In this project we will try to find better ways to assess treatment success by measuring changes in metabolites - substances produced when the body breaks down food, drugs or chemicals, or its own tissue.  This breakdown process is affected in people with acromegaly and our pilot study showed that there are differences in the metabolites in the blood of people with acromegaly in comparison with people without acromegaly.  We will compare changes in metabolites from people with acromegaly who do and do not respond to treatment, and from people without acromegaly, to identify metabolites that can be indicators of whether a patient is responding to treatment and provide clues on how acromegaly is damaging the body.  Thus, this novel research could affect real change in the treatment of acromegaly.

PI: Robert Neely. Co-Is: Anthony Samuel, Dave Smith, Fabian Spill, Sabrina Kombrink, Agi Haines, Paul Roberts
Artist in Residence: Agi Haines

The epigenome can be thought of as a series of switches in our cells. Each switch controls the amount of protein that can be produced in the cell. These proteins are important because they are the machines that manage a cell’s health and its response to disease, environment, ageing and so on. The epigenome is the control panel of the cell. The problem for us as scientists is that the switches on the control panel are invisible.

We have developed a novel technology that will shine a light on these switches in our cells. We can see these epigenomic modifications for the first time and now we need to understand how they are connected and the role they play in the behaviour of a cell. For example, in a cancer cell, we expect different switches to be set to ‘on’ than in a healthy cell; those switches may lead to the production of proteins that make cancer deadly.

We will use this project to unravel some of the unseen wiring between the switch and the cell type by developing new methods for image analysis. This will allow us to rapidly identify healthy and cancerous cells from the array of switches we see and will have application in cancer diagnosis and the development of new cancer therapies in the future.   

PI: Andrew Bagshaw. Co-Is: Caroline Richards, Samuel Johnson, Stefano Seri, Alice Winsor, Leandro Junges, Daniel Galvis
Artist in Residence: Karina Thompson

Epilepsy is a common brain disorder that causes repeated seizures. It can be very difficult to find out if a child has epilepsy, and when they have other conditions at the same time, this can be even harder. Two of the most common conditions that affect children with epilepsy are autism and ADHD. They also affect children's behaviour and brain function. When epilepsy, autism and ADHD are present, it can be very difficult to come to a clear diagnosis and to decide treatment.

In this work, we will use recordings of brain electrical activity from children with and without these conditions to develop computer models of their brains. These models will allow us to understand how each of these disorders affect the brain. In the future, this information will help doctors when they try to assess if a child has these conditions and the best way.

PI: Nicholas McCarthy. Co-Is: Sylvie Freeman, Kamila Silverio Fernandez, Meurig Gallagher
Artist in Residence: Charlotte Dunn

Acute myeloid leukemia (AML) is a life-threatening blood cancer that affects around 3,200 people per year in the UK. Although treatments are improving, AML is still fatal in the majority of cases. 

Treatment for AML consists of chemotherapy, which kills the leukemia. The initial response to treatment is good in 60-80% of patients. However, it is often found that some cancer cells survive in the bone marrow, and these grow in number to cause the cancer to return (‘relapse’) months to years after treatment. 

Highly sensitive blood tests can detect these few surviving cancer cells, called ‘measurable residual disease’ (MRD). Patients with a positive MRD test are more likely to relapse. MRD testing can help doctors decide which treatment is best for patients, and helps test new treatments in clinical trials. 

Testing for AML is highly accurate for diagnosing this type of cancer. After treatment, experienced specialist doctors/scientists are required to analyse and interpret the test results, as it is harder to identify cancer cells at very low levels.  Our research will create software that can analyse MRD results in a consistent manner. This will improve the after-treatment MRD test accuracy and make it easier for the test to be carried out, allowing more people to use MRD testing to improve outcomes for AML patients.

PI: Gerard Cummins. Co-Is: Neel Sharma, Rachel Cooney, Venkata Lahari Balantrapu

Crohn's disease is a condition that causes inflammation of the digestive system (also known as the gut). Inflammation is the body's reaction to injury or irritation and can cause redness, swelling, and pain. Crohn's is a persistent and incurable condition that can affect the patient's quality of life at various points over their lifetime. The long waiting times for endoscopy result in a treatment delay.

This could be tackled through increased capsule endoscopy (CE) usage. This capsule has been in use for 20 years. It consists of a swallowable camera pill that sends data to an external wearable recorder. This technology allows endoscopy to be safely done outside the hospital. The capsules are safely passed out of the body and are flushed away. This technology reduces hospital pressure and potentially reduces the delay between the first signs of illness and confirmation of disease. However, each CE takes time to read and is tedious work, and mistakes can be made, with 6–20% of the signs of disease being missed. Computer programs have been used to help doctors detect gut disease using the more common tube-like endoscopes inserted down the throat or into the bottom.

However, these programs are not suited for CE since the camera is not as good, and the capsule motion cannot be stopped once swallowed, leading to some blurring. This project will use capsule video footage to see how well computer programs can detect Crohn's disease. Crohn's disease was chosen for this project as it affects the lives of 1 in every 650 people in the UK. However, there are few programs developed to aid the detection. 

A computer program that works with the capsules to detect Crohn's can help make it easy for doctors by flagging up specific capsule images for review. Furthermore, it reduces the risk of missed signs of disease and helps speed up the time between the first signs of illness and treatment to manage this disease. 

PI: Deena Gendoo. Co-Is: Andrew Beggs, Chris Lam

PDO’s are tumor samples grown outside the body in clinical conditions in order to create an ‘Avatar’ for testing potential treatments, before the patient undergoes potentially gruelling treatment. To efficiently use these PDO’s, they first need to be checked to make sure they accurately represent the condition of patient.

This project focuses on ensuring that the grown PDO sample matches the patient as closely as possible, in order to be able to develop future treatment. The project focuses on colorectal cancer, where many tumors do not readily respond to drug treatments, and therefore new approaches to understanding the tumor are necessary.

By using new methods of comparing patients and PDOs, this project will help quickly match PDOs to patients, exclude PDOs that don’t exhibit a match, and save time in selecting PDOs for future treatment

PI: Ben Gibbison (University of Bristol). Co-Is: Eder Zavala, Jamie Walker (University of Exeter), Gianni Angelini (University of Bristol), Stafford Lightman (University of Bristol), Daniel Galvis
Artist in Residence: Pietro Bardini

35,000 people a year have heart surgery in the UK. Two-thirds of people make a straightforward recovery and go home quickly. About 1 in 4 people stay in the intensive care unit (ICU) longer than usual. The major cause of this is inflammation - similar to the inflammation that occurs after a sprained ankle. Instead of occurring in a small area, as in a sprained ankle – it occurs across the whole body. Uncontrolled inflammation can lead to failure of the body’s organs. One of the things that protects the body from excessive inflammation is the steroid hormone cortisol, which increases after heart surgery. Doctors sometimes give steroids to patients who are on ICU to prevent severe inflammation. Because we do not know how cortisol is controlled, we cannot produce tests to work out who may need steroids or design the best therapies to reduce inflammation.

We want to model the interactions between inflammation and cortisol. Using data from heart surgery patients, we will then use the model to predict the dynamics of inflammatory responses and look at the differences between people who recover quickly after heart surgery and those who do not. In the future, these models would be useful to inform whether giving steroids at different times and doses could reduce recovery times from surgery.

As part of the project, the team produced this video explaining their work.

This project is funded by the NIHR Bristol Biomedical Research Centre, a partnership between the University of Bristol and University Hospitals Bristol NHS Foundation Trust.

PI: Susan Mollan (University Hospitals Birmingham). Co-Is: Alex Sinclair, Wessel Woldman, John Terry, Leandro Junges
Artist in Residence: Mellissa Fisher

Swelling of the eye nerves (papilloedema) is a frightening diagnosis, which is often found on routine eye examinations in an optician’s shop. Papilloedema is caused by raised brain pressure and can be as a result of serious life-threatening causes such as brain tumours. Sometimes it is difficult for the eye care team to be certain that the swelling is there or not.

Our plan is to use a computer assisted analysis of eye scans and statistical modelling to predict those who have papilloedema and those who do not.

This type of algorithm could have the potential to transform healthcare, by offering earlier and more accurate diagnoses. It would reduce unnecessary emergency admissions by predicting the likelihood of having papilloedema.

PI: David Smith. Co-Is: Zaki Hassan-Smith, Neil Gittoes, Meurig Gallagher
Artist in Residence: Vicky Roden

University of Birmingham research governance reference number: RG_20-088

Hyperthyroidism is a common condition affecting approximately 1 million people in the UK. The thyroids (glands located in the neck) produce too much of certain hormones called ‘T3’ and ‘T4’. This condition can occur for a number of reasons, including growths and the immune system being over-active. If these hormones cannot be controlled properly, they can have serious effects such as heart failure or osteoporosis.

Doctors treat hyperthyroidism with drugs to reduce how much hormone is made. The amount of drug needed is hard to predict, so patients have to return to check and vary their dose, often several times. This is inconvenient and costs a lot of money. It can be bad for the patients if their hormones are too high or too low.

We will solve this problem by developing an app to enable doctors to predict the best dose. The app will take into account the available data: hormone levels at the beginning and early stages of treatment, age, sex, and weight.

The app will then say what is likely to happen to help the doctor decide what to do. The maths inside the app will be based on patient records from Queen Elizabeth Hospital Birmingham.

We will also try out new ways to help patients track their condition when they are not visiting the doctor. An example device is a fitbit – a watch that measures heart rate, sleep and activity. This will help say how well the dose is working for the patient. High heart rate, lack of sleep, over activity then crashing, are signs the dose is too low. Low heart rate, tiredness all the time and lack of activity are signs the dose is too high. All of these symptoms have a big effect on quality of life, and might tell us more than just looking at numbers from blood tests.

PI: Fabian Spill. Co-Is: Melanie Madhani, Daniel Tennant, Peter Ashwin (University of Exeter), Tanja Zerenner (University of Exeter)

Heart attack is one of the major factors of death in the UK and worldwide; yet, we still have no drugs to help survival. Heart attack occurs because the blood supply to the heart muscle is cut off. This causes the heart cells to die due to lack of oxygen and nutrients. Within each heart cell, there are organelles called the “mitochondria”. These mitochondria provide the heart cell with the energy to contract and pump blood around the body in normal conditions. When a heart attack limits oxygen supply, these mitochondria can trigger the death of a cell. Paradoxically, cell death can also be triggered after doctors have resupplied blood to the heart following a heart attack.

Mitochondria are very dynamic, and, for example, can change their shape and their localisation within a heart cell in response to a lack of oxygen. We suspect that this dynamism is related to the triggering of cell death. We are therefore developing mathematical models that can predict how cell death is triggered and use it to uncover how drugs targeting mitochondria may stop cell death. We will also perform microscopy of the mitochondria to investigate how their shape affects drug response.

PI: Peter Tino. Co-Is: Thomas Upton (University of Bristol), Georgina Russell (University of Bristol), Eder Zavala, Krasimira Tsaneva-Atanasova (University of Exeter), Stafford Lightman (University of Bristol), Yuan Shen, Diane Fraser (University of Exeter), Xinyue Chen

Hormones, including the stress hormone cortisol, are released in rhythmic patterns. This means that there are variations in normal levels across time and between individuals. Consequently, traditional, one-off hormone measurements are extremely difficult to interpret, which can lead to delay in diagnosis and treatment.

To address this problem, we need to understand hormone behaviour over the day. We will explicitly model biological mechanisms involved in hormone dynamics. Given hormone measurements taken from patients and healthy volunteers over 24 hours, we will investigate how the data could be explained and classified with the help of our modelling. This will be achieved via the development of a mathematical framework that will enable us to quantify to what degree the data from a particular subject “is explainable” by models related to normal or pathological cases. In other words, to what degree the subject should be diagnosed as a normal or pathological case.

The primary effect will be to benefit patients. Diagnosis will be understandable in terms of the underlying biology, faster, more reliable and convenient. Treatment will be easier to monitor and tailored to the individual. The overall effect will be to reduce the burden on our healthcare system.

PI: Kyle Wedgwood (University of Exeter). Co-Is: David Hodson, Isabella Marinelli, Daniel Galvis
Artist in Residence: Carol Breen

Diabetes is characterised by a loss of control over blood sugar levels. If not properly managed, high blood sugar levels can lead to damage to the heart, kidneys, feet and eyes. One of the key regulators of blood sugar levels is a hormone called insulin. Insulin is made naturally by our bodies by cells in the pancreas called ‘beta cells’. If beta cells do not work properly, they cannot produce enough insulin to keep blood sugar levels under control. Over time, this leads to diabetes.

We often think of beta cells as being the same as one another, since this makes them easier to understand and treat when they become diseased. However, recent studies have shown that differences between beta cells are important for them to work properly. In fact, forcing the beta cells to all be the same can actually accelerate diabetes progression. In this project, we will create a mathematical representation of groups of beta cells and the differences between them. We will use this representation to explore exactly why these differences are important to beta cells. A better understanding of this will allow us to create new ways to treat and prevent diabetes.