Computational and Disease Genomics

Every human carries thousands of DNA spelling differences, but only a tiny fraction of them actually cause disease. This research hunts for those dangerous changes by tracking how proteins bind to DNA—if a single letter change disrupts that binding, it may raise a person’s risk of complex diseases like diabetes or heart disease, or trigger cancer. The problem is that disease-causing DNA changes are needles in a haystack of harmless variation. Most studies start with a suspected disease and look for associated DNA changes. This project flips that approach: it starts with the protein–DNA interaction and lets the human genome data reveal which disease it matters for, without guessing in advance. In cancer, the team compares DNA and RNA from single cells to map how tumours arise and evolve, and whether multiple cancers in the same patient compete or cooperate. If successful, this work could improve how doctors predict disease risk from a person’s genome. It may also reveal new targets for drugs that correct faulty protein–DNA binding. The cancer single-cell work could help clinicians decide whether to treat co-occurring tumours as separate threats or as a single, cooperating system. This is fundamental science—it does not deliver a test or treatment tomorrow, but it builds the interpretive toolkit needed to make sense of the vast amounts of genomic data already being collected in clinics.