Don't miss our weekly PhD newsletter | Sign up now Don't miss our weekly PhD newsletter | Sign up now

  Using big data to better define disease types and predict outcome in childhood myositis.

   GOSH BRC Applied Child Health Informatics Theme (Non-Clinical) PhD Studentships

This project is no longer listed on and may not be available.

Click here to search for PhD studentship opportunities
  Prof Lucy Wedderburn, Prof Mario Cortina Borja, Dr Merry Wilkinson  No more applications being accepted  Funded PhD Project (UK Students Only)

About the Project

Juvenile Dermatomyositis (JDM) is a rare serious childhood autoimmune disease presenting with muscle and skin inflammation, which can affect any organ, and carries burden of serious morbidity and even mortality. Current treatment involves long-term immunosuppression (steroids, methotrexate), but is not targeted, since specific pathological mechanisms of JDM are unknown(1). AT GOSH/ICH we host the largest prospective cohort study of JDM (so-called JDCBS;17 UK centres), with linked longitudinal biological and clinical data (mean follow-up 7.6yrs) on 700 cases of childhood myositis.

Work of Dr Merry Wilkinson (MW, previous BRC Catalyst fellow, proposed co-supervisor) recently identified that in addition to the well-documented interferon signature in JDM, there is a strong signal of dysfunctional mitochondria, detected by transcriptional analysis of immune cells, which is most marked in blood monocytes(2). Our data suggest that oxidised mitochondrial DNA is released and detected intracellularly (via nucleic acid sensing or TLR ligation). Interestingly this signal does not resolve on immunosuppression, even when IFN signature normalises (Wilkinson et al 2023(2)). MW has replicated the MGS signature in a large validation JDM cohort (n=60) and started spatial transcriptome analysis on muscle tissue, from matched patients. We have defined the key leading genes which best represent this Mitochondrial-gene-signature (MGS) and optimised an NCounter (Nanostring) high-throughput assay for measuring this. Pilot results show a high degree of reliability.

There is an urgent need in GOSH/ICH for capacity building in data science: this is an added-value aspect of the project. We address this need by training a biomedical/data scientist in both the omics and clinical data in this project and in current modern statistical and Machine Learning (ML) methods. The student will become expert in analysis of a range of high-dimensional, longitudinal data and will apply methods that are well-suited for this task. The project will cover the two central aspects of modern statistical and machine learning methods, namely estimation and prediction/validation. The student will be trained to an advanced level of programming in the R statistical language, and will benefit from the supervisory team’s links with the Royal Statistical Society, colleagues at PPP and UCL Statistical Science Department, as well as the growing community of spatial tissue biologists at UCL and beyond. All recruited patients, clinical data samples and approvals required for this project are already in place, through JDCBS. Student will be well supported by the JDCBS Study team and wider collaborative myositis scientific network.

Mathematics (25) Medicine (26)

 About the Project