Dr Dowsey's research concerns signal/image analysis and statistical modelling applied primarily to post-genomic data. In particular, the goal is to provide a fully automated discovery engine for alignment, quantification and modelling of cross-omics data, whether by mass spectrometry, liquid chromatography, 2-D gel electrophoresis or any emerging technique. The Bayesian modelling is powerful but kept quite general and flexible towards the ultimate aim of linking cross-scale comparison and pathways and analysis across omics datasets (transcriptomics/proteomics/metabolomics). To this end, accelerated computing techniques such as GPU and Grid infrastructure are key.
A novel and general pipeline for sensitive analysis of statistically sparse datasets, with emphasis on mass spectrometry and surrounding techniques. In particular, a fundamentally new paradigm for feature detection in mass spectra is presented which is able to detect features barely discernible from noise and separate overlapping ones. The 'seaMS' framework opens up a wide range of research directions for further integrated modelling across and within raw omics datasets.
The BMSS poster contains the most recent summary of seaMass, including new quantification results. The BSPR talk transcript provides a general introduction.
Personal details | Research | Publications
This website will look much better in a web browser that supports web standards, but it is accessible to any browser or Internet device.