* Volcano plot provided by Dr. Simon-Pierre Gravel and Dr. Gaëlle Bridon

Due to the time commitment of untargeted analysis, the MCF currently only offers this service to GCRC members.  We hope to offer this service more broadly in the near future.

Untargeted analysis will be managed as a close collaboration between the MCF and the investigator.

Three instruments can be used for untargeted analysis: LC/QTOF, GC/MS and NMR.

A general workflow is described below. This workflow is flexible and will be modified to meet the needs of the project.





General workflow:

Samples are extracted according to the chemistry of the metabolites that the scientist is most interested in (e.g. polar metabolites vs fatty acids).

  • NMR: Samples are prepared and data are acquired 
  • GC/MS: samples are derivatized (usually MOX/TMS or MOX/TBDMS) and GC/MS data collected
  • LC/MS: an appropriate column and buffer conditions are chosen depending on the class of molecules of interest (e.g. organic acid vs nucleotide) and LC/QTOF data are acquired.


  • NMR data are fit using Chenomx software or other software as defined by the project.
  • GC/MS data are deconvoluted by chromatography and retention time using AMDIS or MassHunter Qualitative software (Agilent).
  • LC/QTOF data are processed for molecular features including all ions (plus their isotopes) and adducts. This may be repeated in a recursive workflow using MassHunter Qualitative software and Mass Profiler Professional (MPP) (Agilent).



  • NMR assigned metabolites, bins, or modeled peaks are subjected to statistical analysis (PCA, PLSDA, etc) using MPP (Agilent) or MetaboAnalyst.  Metabolites or peaks causing the differentiation between groups are identified.
  • GC/MS spectra are tentatively identified using Feihn, Bains (Steadman Metabolism laboratory, Duke University) or NIST11 databases. The identified and unidentified metabolites are subjected to data alignment across different samples and statistical analysis using MPP. Identified and unidentified metabolites that cause differentiation between groups are “tagged” for further identification. 
  • LC/QTOF extracted metabolites or “features” are subjected to data alignment across different samples and statistical analysis using MPP. Differentiating “features” are “tagged” for identification.


  • NMR : metabolites identified using 1D and 2D NMR techniques. Significant metabolites may be spiked into samples to confirm chemical shift, spectral patterns and a quantitative increase in the metabolite’s concentration. The HMDB , MMSD and BMRDB are used to help identify metabolites and their spectra.
  • GC/MS based metabolites are identified and ideally new samples are split where one is spiked with the authentic metabolite standard(s) confirming an increase in the putatively identified metabolite and confirmation of mass spectrum and retention time.
  • LC/MS: samples are re-run where the statistically significant differentiating metabolites are targeted for MS/MS analysis using different collision energies (Product ion scanning). These data are then used to identify the metabolites (tentative). The METLIN accurate mass and MS/MS database as well as the Human Metabolome Database are used to help the identification. If possible, samples are spiked with authentic metabolite standards to confirm an increase in the significant metabolite peak intensity as well as to confirm retention time, mass spectrum and fragmentation pattern.

Once metabolites have been identified, the pathways where they are up-regulated and down-regulated need to be elucidated. Using Pathway Analyst (Agilent), Cytoscape (NRNB), MetPA (University of Alberta) etc, can facilitate the identification of active pathways. This data can be compared to other «omics» data (transcription, protein levels etc). Further analyses such as pathway inhibitors, genetic manipulations and flux analysis are encouraged to confirm pathway use and directionality.