Why should I use a reference metabolome?

Currently, no single analytical technique can simultaneously measure all the metabolites involved in the metabolic pathways. Different platforms - NMR, GC-MS, LC-MS, usually have different compound coverage. These difference will likely introduce bias during the enrichment analysis, which assumes all compounds in the pathways or metabolite sets have equal chance to be measured. To correct for this bias, a reference metabolome that is specific for the platform is recommended. Please note the reference metabolome should be uploaded as a list of common compound names.

Please note, this correction is only necessary when only a list of significant compounds are provided (over-representation analysis). When the whole concentration table is provided, the compounds within the data are used as the reference metabolome. It is not necessary to upload separately a compound list as reference metabolome.

How is the custom reference metabolome (if provided by the user) used by MetaboAnalyst?

  • Is it in order to perform accurate correction for multiple testing?
  • Quantitative Enrichment Analysis (QEA) typically uses as input all detected metabolites. Does this mean that for QEA we practically never need to provide a custom reference?