If you need to set-up and validate the potential of LC-MS for non-targeted metabolomics, Progenesis CoMet makes it much simpler and quicker. Our application note shows data analysis could be performed in hours, not weeks, to demonstrate LC-MS would differentiate pig adipose samples with and without “taint”, a characteristic that impacts boar meat production.
In many countries male piglets are castrated shortly after birth to avoid the production of meat with an unpleasant smell and flavour known as boar taint. This has financial and ethical issues in terms of raising animals and providing good quality food. Several compounds have been reported to be associated with this condition, however, the level of these compounds does not always correlate with results from classical sensory panels and other factors are thought to be involved1.
The application note was written with help from researchers in Zurich, Switzerland at the Institute of Veterinary Pharmacology and Toxicology and the Functional Genomics Center. Specifically, using data provided by Malin Olson from her PhD project titled “Multiplex Profiling of Boar Taint by Non-Targeted Metabolomics”1.
The result of this original research demonstrated the feasibility of LC-MS to discriminate tainted and non-tainted carcasses based on a sub-set of biomarker compounds. However, some of the challenges that had to be resolved over months of work included:
- limited access to LC-MS system
- applying many separate applications
- specialist biostatistical support for interpreting results
Our data analysis solution
Over a hundred LC-MS runs generated from a nanoAcquity UHPLC® connected to a Synapt G2™ HDMS™ mass spectrometer were reanalysed by Progenesis CoMet. Our aim was to see if the same proof-of-concept results could be produced using a much simpler approach. Our approach also included several restrictions compared to the original analysis, including:
- we wanted to complete analysis and review results within a working day
- we were not able to run the same compound database searches as the original study
- we used the univariate and multivariate statistics built into our software rather than use complex off-line applications or rely on input from biostatisticians
You can download the application note for details of the method and results. But in summary we automatically generated compound quantification and identification results from Progenesis CoMet using default parameters. The first check on the results was a PCA plot based on a list of compounds that showed a “significant” (p<0.05) abundance increase in tainted samples relative to non-tainted samples.
Compound identifications, or RT and m/z measures for unidentified compounds, were compared between our list of eighteen significant compounds and a final list of sixteen compound biomarkers chosen in the original research1. Ten of the eighteen compounds were found to be common, three with identifications confirmed as testosterone, androstenadione and 3-oxohexadeanoic acid. In addition, nine new compounds of interest were found compared to the existing list of sixteen compound biomarkers.
Progenesis CoMet showed advantages in simplifying and speeding up data analysis, providing comparable proof-of-concept results in less than 3 hours. This can be a great benefit in providing speed, objectivity and accessibility of running discovery-focussed experiments prior to committing further resource into research.
“Untargeted metabolomics, where there is no known answer, is a challenge that can involve applying many different applications and significant time to generate results. Progenesis CoMet provides a single workflow combining all the major steps needed for semi-quantitation and annotation of putative compounds for further validation that can quickly validate LC-MS approaches.”
Malin Olson, Institute of Veterinary Pharmacology and Toxicology, Zurich, Switzerland
1. Olson M, Laczko E, Lewis F, Ampuero S, Bee G, Naegeli H. Multiplex Profiling of Boar Taint by Non-targeted Metabolomics. 2012.