Project X: Statistical
methods in spectroscopy in genomics, proteomics and metabonomics.
Mass spectrometry
is a commonly used technique in proteomics, but spectroscopic methods in
general are used in bioinformatics; a recent study used statistical methods for
classification of tissue samples in cancer studies. Data reduction is a key aspect, and this will
be the focus of the project.
Fortunately, the computational aspects have been studied extensively,
and many well established techniques have been developed.
Similar techniques are also used in the
rapidly developing field of metabonomics. High frequency proton NMR spectroscopy
provides a rapid method of characterizing and quantifying a wide range of
metabolites in untreated biological fluids and a powerful means of exploring
the biochemical consequences of disease and toxicological processes. The
coupling of biofluid NMR spectroscopy to mathematical data reduction methods
give us a new metabonomic
approach to disease diagnosis complementary to proteomics and genomics and can
readily be applied in vivo in essentially non-invasive studies.
In this project, different statistical and
computational approaches to, for example, data reduction and extraction will be
studied.