1. Metabolomics and the Human Metabolome Project
David Wishart & Lori Querengesser, University of Alberta, Edmonton, Canada
Abstract: Unlike the situation in genomics, where the human genome is now fully sequenced and freely accessible, metabolomics is not nearly as developed. It is estimated that only ¼ to ½ of endogenous human metabolites in blood or urine have been positively identified. Of those that have been identified, very few have any information on their normal concentration ranges. Further, there is no publicly available, electronic database for human metabolites (i.e. no Metabo-Bank) and there are no “metabolite libraries” which allow researcher to obtain quantities of rare metabolites from which to standardize their instruments. As part of our newly funded Genome Canada project we are working towards completing the human metabolome. This will involve identifying, characterizing, and quantifying an estimated 1400 endogenous metabolites that can be found in urine, blood, CSF and white blood cells at concentrations greater than 1 micromolar. The data and the compounds isolated, synthesized or purchased from this work will be used to create a freely available electronic database – called the human metabolite database (HMDB) and an open access chemical warehouse – called the human metabolite library (HML) with mg-gram quantities of these metabolites.
Introduction
Project Workflow
The HMDB
Metabolomics is a newborn cousin to genomics and proteomics. Specifically, metabolomics involves the rapid, high throughput characterization of the small molecule metabolites found in an organism. Since the metabolome is closely tied to an organism’s genotype, its physiology and its environment (what the organism eats or breathes), metabolomics offers a unique opportunity to look at genotype-phenotype as well as genotype-environment relationships. Metabolomics is increasingly being used in a variety of health applications including pharmacology, pre-clinical drug trials, toxicology, transplant monitoring, newborn screening and clinical chemistry. However, a key limitation to metabolomics is the fact that the human metabolome is not at all well characterized.
project we are working towards completing the human metabolome. This will involve identifying and quantifying an estimated 1400 endogenous metabolites that can be found in urine, blood, CSF and white blood cells at concentrations greater than 1 micromolar. The data and the compounds isolated, synthesized or purchased from this work will be used to create a freely available electronic database – called the human metabolite database (HMDB) and an open access warehouse of chemicals – called the human metabolite library (HML).
A project of this type requires considerable coordination between mutliple labs. Using models originally developed for the Human Genome Project we have devised the following workflow.
Over the next 2.5 years we expect to compile detailed data (spectra, phys-chem parameters, and concentration data) on most kinds of human metabolites. Already we have compiled data on 200 metabolites and placed them into a first draft of an on-line human metabolite database (HMD). It is available at
To date 300 chemical compounds have been purchased or acquired and another 400 have been identified as the next to be acquired, synthesized or isolated. Approximately 100 new compounds are added to the database every month.
Patient
Pool
Prelim Bio. Sample
Acquisition, prep and
GC/MS characterization
F. Bamforth
Ethics Board
Approvals
Compound ID
Lit Search
Data Mining
Greiner/Wishart
Compound
Purchase and
Archiving
Wishart
HMDB web page
LIMS, Sys Admin
Wishart/Greiner
Project Admin &
Coordination
Wishart/PM/AA
Characterization
NMR
Wishart/Sykes
FT-ICR MS
Ellison/Li
Quantitation
Vogel
Further Cmpd
Separation
And Isolation
Wishart/Li/Ellison
Compound ID &
NMR/MS
Wishart/Sykes/Li
Vogel/Ellison
Synthesis
Clive
1.5 FTE
2 FTE
1 FTE
5 FTE
3 FTE
Project Participants
The Human Metabolome Project brings together 8 PI’s from a broad variety of backgrounds including clinical chemistry, biochemistry, analytical chemistry, organic chemistry & computing science. It is expected that another 20 individuals will be hired to work on various aspects of the project. We are hiring now!
Metabolomics Tools
Metabolomics
Proteomics
Genomics
High throughput metabolic profiling or metabolomics involves using either NMR, FT-MS or HPLC methods. We will be using all 3 kinds of instruments to characterize 3 different kinds of compounds: 1) purchased 2) synthesized & 3) isolated from biofluids
1400??
Chemicals
3000 Enzymes
30,000 Genes
Brian Sykes
Biochemistry
U of Alberta
NMR spect.
Russ Greiner
Comp. Sci.
U of Alberta
Bioinformatics
David Wishart
Comp. Sci.
U of Alberta
Proj. Leader
Hans Vogel
Biochemistry
U of Calgary
NMR spect.
Figure 1. The Pyramid of Life, a diagram illustating the relationship between metabolomics, genomics and proteomics.
A Modest Proposal
Figure 2. Screen shot montage of the Human Metabolome Database web pages. These shots show examples of the browser/sorter (top), the “metabocard” for each metabolite (50+ fields of data – left), the structure viewing applet (right) and the structure query tool to allow individuals to search by structure similarity (bottom). All metabolites will be linked to known biochemical pathway diagrams and enzyme or gene data.
It is estimated that only ¼ to ½ of endogenous human metabolites in blood or urine have been positively identified. Of those that have been identified, very few have any information on their normal concentration ranges. As part of our newly funded Genome Canada
Fiona Bamforth
Clin. Chemistry
U of Alberta
Sample Acq.
Derrick Clive
Chemistry
U of Alberta
Synthesis
Liang Li
Chemistry.
U of Alberta
MS/Separation
Mike Ellison
Biochemistry
U of Alberta
MS/Separation.