My main research interests focus on identifying varying types of (weak) signal in crystallographic datasets. Through the application of statistical methods to multiple datasets, I develop approaches to increase the amount of signal that can be extracted from the data relative to conventional single-dataset methods. These methods enable a more confident interpretation of the (otherwise noisy) crystallographic data.

Previous projects have focussed on the identification of ligand binding in the context of fragment screening experiments: identifying when ligands binding occurs, but also identifying any conformational changes in the protein that happen as a result.

More recently, I've started work on the identification and modelling of heterogeneity in protein crystals; I'm now developing methods to identify and model any heterogeneity that is present. I also work on methods for validating the models of weak crystallographic features.