We develop novel bioinformatics methods to analyse and make sense of cancer '-omics' data. A major focus of our work is cancer evolution and intra-tumoral heterogeneity, particularly in melanoma. One of our aims is to develop practical measures of tumour heterogeneity and test if these are predictive of patient outcome.
We are also interested in complex genomic rearrangements and the mechanisms underlying extreme amplification events in genomes. We have developed novel methods to identify these rearrangements and use mathematical models to make sense of them. With this approach we recently discovered the dynamic mechanisms underlying the formation of giant cancer-associated neochromosomes.