Transposable elements in cancer

We develop and apply integrative methods to measure how much aberrant gene expression in human cancers is driven by transposable elements, and to pin down the regulatory and epigenetic pathways that allow it.

• Which elements are reactivated, and in which tumours?
• Which chromatin and DNA methylation changes release them?
• How do they rewire nearby genes?

Clonal haematopoiesis and inflammation

Clonal haematopoiesis, where blood stem cells carrying a mutation expand over time, is linked to blood cancers and cardiovascular disease. We found that the common DNMT3A and TET2 mutations are associated with distinct patterns of retrotransposable element activity and inflammatory signalling, pointing to different routes to disease.

Predicting treatment response in AML

Doctors still cannot reliably tell which patients with acute myeloid leukaemia will respond to chemotherapy. Using data from the Beat AML cohort, we built multilayer machine learning models and showed that adding endogenous retrovirus activity improved prediction of induction chemotherapy failure.

New targets for immunotherapy

When transposable elements are spliced into genes, the “exonised” sequences can produce new peptides. Some may be recognised by T cells, and others add new structures to cell surface proteins that antibodies could target. We look for these as candidates for highly selective therapies.

Methods for multi-omics data

Underpinning all of this is method development: tools to quantify repetitive sequences, integrate multiple data types, and explore them visually so that biological signals are easier to spot and test.