|
|
 |
|
| Metabolomics Science |
|
|
Showing posts for the Article of Interest - Koulman et al. Anal Bioanal Chem. 2009 March 11; [Epub] topic:
| AUTHOR |
POST |
dmutch@uoguelph.ca
[Apr 15 2009 at 12:02:13 PM] |
As the field of metabolomics continues to build momentum within the scientific community, Koulman and colleagues have provided a timely reminder that the identification of biomarkers is a long and arduous process. Many researchers (and I’m guilty of this myself) publish studies describing metabolite differences between two populations (e.g. diseased vs. healthy) and then conclude that they have identified biomarkers with possible clinical relevance. While this certainly may be very true, the required follow-up work to confirm its potential as a biomarker is rarely performed. This is partly because of the time-consuming nature and complexity of such validation studies. Koulman et al. have suggested an analytical pipeline that outlines the steps required in order to fully exploit the promise of metabolomics research for the identification of biomarkers. Their pipeline is divided into 4 phases:
1) Discovery Phase, which focuses on identifying those metabolites that differentiate two states (e.g. disease vs. healthy, before and after a nutritional or pharmaceutical intervention, etc.). Metabolites identified at this stage should be considered ‘differentiating metabolites’.
2) Validation 1 Phase, which will focus primarily on the replication of Discovery Phase results. In other words, if one examines an independent population, does the differentiating metabolite continue to differentiate the two groups of interest?
3) Validation 2 Phase, will examine the relationship between the differentiating metabolite and a clinical trait in order to determine the specificity of this candidate biomarker. For example, as the degree of severity of a disease worsens, levels of the candidate biomarker change accordingly. Furthermore, additional confounders such as age, gender, diet, physical exercise, other related disease states, etc must be shown to have little to no effect on the association between the candidate biomarker and the clinical trait of interest.
4) Application Phase, where the candidate biomarker will be compared to existing markers in order to demonstrate some improvement over the current diagnostic tests. Furthermore, a practical and rapid diagnostic test must be developed.
While the above pipeline is certainly daunting, it is by no means a reason to “throw the towel in”. Indeed, the field of metabolomics is poised to have a tremendous impact on the management of health and disease. Discovering metabolites that improve our ability to predict, monitor, and treat a disease compared to existing methods is a realistic goal; however, to accomplish this goal means that the research community must continue beyond the simple step of identifying differentiating metabolites.
Full Article: Koulman et al. Anal Bioanal Chem. 2009 March 11; [Epub]: PMID: 19277615.
|
Back to Topic List
|
|
|
|
|
