A blood test now predicts the long-term course of inflammatory bowel disease, and researchers want to build similar tools for multiple sclerosis, lupus, and other immune disorders. Immune diseases such as MS and Crohn’s affect up to 10% of people at some point in their lives. They are incurable, often strike in young adulthood, and vary wildly between patients—yet doctors have few reliable ways to forecast who will fare well and who will deteriorate. The team has spent nearly two decades collecting detailed patient data, including immune cell counts, blood protein levels, and disease activity scores. Their previous work turned this data into a test that helps IBD patients understand their likely disease trajectory and choose treatments accordingly. Now they plan to use artificial intelligence to analyse across multiple diseases simultaneously, drawing strength from the complexity of the data. They have also added a cohort of severe COVID-19 patients, over 10% of whom show signs of autoimmunity. The goal is to produce new predictive tests for each condition, along with publicly available data and analysis pipelines for other researchers. If successful, this would move immune disease management toward personalised medicine—tailoring treatment to the individual rather than applying a one-size-fits-all approach.
View original technical description
Diseases caused by a malfunction of the immune system are common and can be severe. They include multiple sclerosis (MS) and inflammatory bowel disease (IBD: Crohn's and ulcerative colitis), among many others. They are incurable, often occur in young adulthood, and effect up to 10% of the population at some time in their lives. We are now able to analyse samples from patients with these diseases, and produce datasets of unprecedented detail. Our challenge is to use this information to increase our understanding of disease, and to deliver tests to help guide patient treatment and improve outcome. Our previous MRC programme grant had begun this process through analysis of patient data collected since 2004, which led to a test which predicts long-term outcome in IBD - this is now available to patients to inform them of their likely disease behaviour and to help guide treatment. This grant also funded the development of new, enlarged patient cohorts encompassing different immune diseases: MS, IBD (both Crohn's and UC), SLE, ANCA-associated vasculitis and idiopathic pulmonary fibrosis. We now propose to exploit this newly created resource by developing "integrative analysis" methods to make sense of the data we have collected. These use AI (artificial intelligence, or machine learning) to analyse across different sorts of data (including things such as immune cell numbers, blood protein levels and disease activity scores) and different diseases over time, drawing strength from the complexity of the data collected. In addition to the diseases listed above, for comparison we have also included a cohort of patients with severe COVID-19, all of whom have severe inflammation and over 10% of whom also have autoimmunity (evidence of the immune system directly targeting the patient). This promises to lead to new insights into what drives disease, and importantly also to the development of further tests to help predict disease outcome and help guide patient treatment. This work will produce publicly available data and new analysis pipelines to support future research by the scientific community, will train young researchers in cutting edge methods, but should also tests of practical value to patients in the clinic - hoping to achieve "personalised medicine" by allowing patient management to be tailored to individual patient needs.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know