Research
Threads that run through the lab — each a different angle on the same question: what can we learn, and predict, from biological data?
Sequence- and structure-based prediction of binding sites and the energetics of complex formation.
Learn more →Genome-wide dictionaries of DNA conformation — twist, roll, slide, groove width — for indirect-readout models of recognition.
Learn more →Novel architectures and training strategies, including language-model encodings, purpose-built for biological problems.
Learn more →Integrating expression, phosphoproteomics and toxicogenomics to translate signals across species toward drug discovery.
Learn more →Deep learning on whole-slide images and RNA-Seq for cancer diagnostics and early detection.
Learn more →AI/ML modelling of antimicrobial resistance and host–pathogen interactions, with a focus on tuberculosis.
Learn more →Harmonizing single-cell and spatial data, dimensionality reduction, and predictive modelling of cell state.
Learn more →Single-nucleus transcriptomics of the human brain to uncover molecular signatures of neurological function and disease.
Learn more →Projects
Ongoing
A genome-wide analysis of the conformational dynamics of transcription-factor targets, building shape-based models of transcriptional recognition.
Read more → OngoingA machine-learning method to identify previously unannotated DNA-binding proteins by integrating sequence and expression evidence.
Read more → OngoingLinking documented genetic diversity across sequenced bacterial and viral genomes to phenotype, including therapeutic-target discovery.
Read more →Recently completed