A review of Metabolomics, Brisbane 2017

After 25 hours of travel… a short rest with some adorable baby kangaroos and koalas, there was a series of pre-conference workshops on Sunday and Monday morning.

Image of Brisbane Bridge at night time

The Waters Monday workshop was well attended, and we had two very interesting presentations from Progenesis users in the FoodOmics and Health Sciences applications: Dr. Martin Snel from South Australia Medical Health Research Institute (SAMHRI) presented “Are identical mice the same? The gut microbiome and fecal metabolome in mouse models” in which he discussed the influence of the microbiome on the phenotype of “identical” mice. Dr. Fe Calingacion (University of Queensland) presented “A multi-omics approach to understand grain quality using a diverse set of rice” in which she discussed how omics profiling may be used to distinguish different types of rice and assess the quality.

At the conference itself, discussions were held on various applications and topics during the sessions.  These included Natural Products, Lipids, Marine Microbiomes, Diet, Health and Disease, Wine and Quantitative Metabolomics. Across this broad diversity of Metabolomics applications, it was clear that the main challenge people are seeing is compound annotation.

Image of poster entitled “Automatic CCS and MS/MS Library Creation and Application for Large Scale Metabolic and Lipidomic Profiling”

Although it is clear that metabolites annotation is still perceived to be the main challenge for the community, missing values (in statistical data for relative quantitative analysis) is also a hot topic. Indeed, from various discussions I had with scientists, missing values are a well-known problem but no one has managed to resolve them in a satisfactory manner so far.

But did you know missing values need not be a critical issue? Progenesis QI software has a solution for the problem in the Co-detection approach (see also How Progenesis QI resolves the problem of missing values). This is vitally important since, in addition to the issues of compound identification mentioned above, it’s also vital that you attempt to identify the correct compounds, i.e. ones that are showing some interesting expression changes in your experiment. Only with confident quantitative results can you be sure that the compounds you found are the potential biomarkers you were looking for.

Another main topic highlighted during the conference was the constant need for better data quality.

I spoke to a bio-informatican whose approach was to use only data with a SD <3 as valid compound ions for analysis and further data exploration – a form of “data cleansing” approach. With the great hardware technologies available nowadays, High-Resolution Mass spectrometers coupled with Liquid Chromatography separation, you can generate large amounts of complex data fairly rapidly. So you need reliable tools to “separate the wheat from the chaff”, enabling valid results and a relevant biological interpretation. Progenesis QI provides a range of QC tools such as the ability to quickly and easily filter the data by CV%. Using “tags”, you can then easily choose to hide or display data based on parameters such as p-value, fold change, compound abundance and highest mean condition, enabling very flexible data exploration and reliable validation of results.

If you’d like to see how Progenesis QI can improve your discovery metabolomics and lipidomics analysis results, you can download the full version from our website along with some tutorial data and a user guide.

Finally, just to encourage you, here is a quote from one of our users:

“We have been running samples for 2 months non-stop, resulting in a data set of 1115 samples, each sample consisting of a 30 minutes gradient separation and an average of about at least 3000 compound ions per sample. A gradually increasing retention time shift up to more than one minute was observed towards the end of the gradient for the last 400 samples, but with Progenesis QI we could align all of them without any exception! This was a challenging task which to my opinion no other software would be able to handle!”

Geert Goeminne
Mass Spectrometry Expert, Department of Plant Systems Biology, VIB, Ghent University, Belgium