Manchester BRC Biomarkers week – Biostatistics
Manchester is a leading centre for biomarker development, brought together by the NIHR Manchester BRC’s biomarkers ‘cross-cutting’ research themes.
Our biomarkers research spans the fields of genomics, proteomics, clinical imaging, and biostatistics, and supports our mission to drive health improvements, and bridge the gap between new discoveries and individualised care.
What are biostatistics?
Biostatistics (short for biological statistics) is the use of statistical tools and tests, often applied to large and complex data sets, to help answer biological questions in research.
A Biostatistician (also called a Bioinformatician) looks at data across different biomarker platforms, such as genes, proteins, or any other bio-molecules, and uses statistical methods to handle, manipulate, compare, analyse and visualise the data, and help answer questions such as:
- which set of molecules, out of the hundreds or thousands measured, can help identify patients with a specific disease?
- what are the clinical characteristics associated with a favourable response to treatment?
- can we identify different groups of patients who will react differently to a specific drug?
Why are they so important?
Clinical research has come a long way from observing symptoms in patients. We’re now able to collect multiple biomarkers and other data about our patients, throughout the length of their routine care, and during clinical research studies.
As we collect more and more data on patients, modern computing means we can quickly compare and analyse large data sets, or even train computers through artificial intelligence (AI) to learn from them and look for things humans may not spot, called machine learning.
Biostatistics can also look at data on patients with the same disease, and break these down into smaller groups that share similar clinical characteristics, such as age, type of drug or treatment, whether they suffer from other conditions, and more. This process is called stratification.
Dr Nophar Geifman, Senior Lecturer in Biomedical Data Analysis and Modelling at The University of Manchester, has supported a number of biostatistics projects across the BRC. She explained:
“Biostatisticians and bioinformaticians support researchers by allowing them to fully maximise the information and knowledge that can be extracted from their data sets. In essence, they help them to make sense of the complex data we can now generate, turning it into information that backs the development of new tools to help patients. These can be anything from an algorithm that predicts risk of developing a condition, to a blood test that can help make a diagnosis or monitor disease progression and response to treatment.
“The BRC’s biomarker theme cuts across all of our clinical research themes. This encourages our theme leads and clinical academics to collaborate with researchers in other specialties, including biostatistics and informatics, and means we can use our expertise to support their research and clinical decision making.
“As we’re able to collect more data on our patients, such as clinical, genotyping, and proteomics, and across various disease areas, these approaches will allow for better patient stratification, biomarker discovery, and even better personalisation and precision of treatments.”
Our biostatistics research – psoriasis care
Psoriasis is a common inflammatory skin disease that causes red, flaky, crusty patches of skin covered with silvery scales, most commonly on the elbows, knees, scalp and lower back.
The disease affects around two million people in the UK, and is a long-lasting (chronic) condition that is currently incurable. Some people may only have mild symptoms, but for others it can have a massive impact on their quality of life.
Biomarkers like this mean we’re moving away from the current trial and error approach to prescribing medicines, to a precision medicine one, tailored to each patient.
Psoriasis is associated with arthritis, heart disease, diabetes and depression. It’s also common for the disease to fluctuate in severity, with people experiencing periods of none or only mild symptoms, followed by more severe symptoms.
The Psoriasis Stratification to Optimise Relevant Therapy (PSORT) study, led by our Dermatology Theme Lead, Professor Chris Griffiths, aims to better understand how patients will respond to biologic therapies. Biologics are a family of drugs used to treat psoriasis. They are given to patients as injections, and reduce skin inflammation by targeting overactive immune cells in the body, which are a cause of the condition.
Part of the study aims to develop a computer algorithm which can look at thousands of patients’ data, including: which biologic drugs they are treated with, the dose and when they started it, and whether they experience skin genetic changes during treatment, and/or changes in the number of immune cells (called T-cells) in their skin and blood.
The algorithm looks for hidden trends across these areas, to help predict how patients will respond to treatments in the future, and which treatment will be best for them.
Professor Griffiths said:
“Manchester BRC’s dermatology research is helping us better understand skin conditions, and drive discoveries into new diagnostics and therapies that could help thousands of people.”
“Patients with psoriasis can respond very differently to biologic therapies. For some patients, biologics can be highly effective for many years. In others, they work well at first but then reduce in effectiveness, whilst for a minority, they are ineffective. Biologics are also expensive, so it’s vital we improve their effectiveness, and get patients on the right biologic for them, first time.
“Using biostatistics and informatics to support our research means we use large amounts of patient data to stratify them by clinical, genetic and immunological characteristics. This allows us to personalise treatment and identify the therapy that is most likely to work before they start treatment. Or if a patient doesn’t respond, we can find out quickly and change treatments.
“Biomarkers like this mean we’re moving away from the current trial and error approach to prescribing medicines, to a precision medicine one, tailored to each patient.”