My name is David Wishart,
I'm with the Departments of Biological Sciences and
Computing Science at the University of Alberta in Canada.
Today, my talk is on "Metabolomics Resources".
So, there are two routes to metabolomics.
One is quantitative or targeted approach,
the second is called untargeted metabolomics or profiling.
What we're showing here is an NMR spectrum of urine,
and with quantitative methods,
typically one is identifying and quantifying all of
the peaks before following on with multivariate statistics.
In untargeted profiling, typically what you do is multivariate statistics first,
which essentially simplifies the spectra and allows you to
identify important features which is shown with these PCA plots,
and then after those important features are identified,
then you can proceed to doing quantification and identification.
So, in essence with metabolomics,
there's a general workflow that almost every researcher follows.
The first part is to collect spectra,
make the NMR, or GC-MS,
or LC-MS/MS spectra, and from there you might
generate long lists of metabolites or peaks.
In this case, we're trying to focus on
targeted metabolomic so typically you'll get a long list of metabolites.
Second, from that long list of metabolites,
you try and use multivariate statistics to
identify smaller numbers of significant metabolites.
Those things have changed,
they've gone up or down depending on the experimental conditions you are looking at.
From that smaller list of significant metabolites,
then you try and move towards pathways,
biomarkers, and biological interpretation.
This is of course the essence of metabolomics and
what mostly we'll try and focus on today.
Once you've done all of that,
you try and deposit your metabolomic data and some of
your interpretation in various public databases.
Today, I'll talk about all four steps in metabolomics workflow.