What is a Standard

1
Metabolomics Ontologies and Standards David Block
and Dan Tzur, University of Alberta, Edmonton,
Canada
Standards, Metadata, and Ontologies
Abstract
Ontologies and Standards Will Be Required For
A remarkable opportunity exists to develop
standards and ontologies for metabolomics based
on end-user needs. While we are developing tools
for collecting, analyzing, and reporting
metabolomic data, interoperability between
different groups can be achieved through the
development of community standards and shared
ontologies, enabling reuse of experimental
results and comparison between locations and
platforms. Common vocabularies aid the
development of data repositories and analysis
tools that can operate with multiple data sources.
Clinician/Scientist Sample handling Sample
pre-treatment Experimental design
Instrumental protocol Units of measurement
Labeling protocol External controls Data
normalization Data editing and analysis
Patient Demographics Phenotype, condition
Description of symptoms Forms for clinical data
collection Sample collection procedure
Patient identification/anonymity
Standards are crucial in defining minimum quality
control requirements for data acquisition and
interpretation. In complex laboratory and
clinical environments, standards serve to reduce
the number of sources of variability, and enable
the comparison of results from different
locations. Improper quality control at any step
in the workflow can result in missing data, and
in the creation of spurious, misleading
information. Including metadata, or data about
the data, allows for the interpretation of
primary results by researchers and clinicians who
were not present at the time of acquisition.
Metadata also allows the grouping of experimental
results that were not originally related.
Standards development is occurring through the
Metabolomics Society (www.metabolomicssociety.org)
. The identification and description of
components of experiments, phenotypes, and
processes relies upon language, which is
imprecise. Using a shared ontology allows
precise identification and reasoning about
related concepts. Many relevant ontologies are
available and being developed at the Open
Biomedical Ontologies website (obo.sourceforge.net
).
Idealized metabolomic workflow A patient presents
with multiple symptoms. A clinician records the
patients demographic data and phenotype on a
standard form, using a controlled vocabulary
defined as an open standard. A sample is taken
according to a defined protocol, and then
analyzed on multiple instruments using published
standard operating procedures. All the
experimental data and metadata are automatically
captured by the instruments and included in all
further communication and analysis. The sample
data is uploaded to the HMDB, where it is
compared to data in other repositories that
report their data using a common format. A new
analysis algorithm developed in another location
can be applied to the data, since it is based on
this common data format. Several analytic tools
in multiple locations are used to analyze the
data, and their results are easily compared,
since they also output in a common format. The
consensus, along with a description of the
complete workflow, is returned to the clinician,
allowing a diagnosis. Follow-up data is
submitted on the results of the ensuing
treatment, validating the hypotheses generated by
the analysis.
Introduction
Metabolomics research requires collaboration
between biologists, chemists, clinicians, and
computing scientists. Researchers need to
describe experiments, normalize, analyze, and
integrate data, and exchange results.
Communication between these disciplines requires
a common language and notation Standard terms
for ideas and concepts kinase vs.
AMP-activated protein kinase complex
GO0031588 Standard formats for message
exchange free text vs. ArMet-compliant XML
(www.armet.org) These standards are in
development. HMDB (the Human Metabolome Data
Base) will support this effort, and adopt these
standards as they are published.
Database Compound identification Phenotype
identification Data normalization Data
analysis
HMDB
Communication File format Semantic content
Platform agnostic documents Required amount of
metadata

• What is a Standard?
• An acknowledged measure of comparison for
  quantitative or qualitative value a criterion
  (www.answers.com)
• reporting standards will specify the data
  identified as necessary for complete and
  comprehensive reporting in a range of identified
  contexts, such as submission to academic
  journals. Data exchange standards will be
  developed to provide a technical vehicle which
  meets or exceeds the requirements of reporting
  standards.(www.metabolomicssociety.org/mstandards
  .html)
• What is an Ontology?
• A description (like a formal specification of a
  program) of the concepts and relationships that
  can exist for an agent or a community of agents.
  In biomedicine, such ontologies typically specify
  the meanings and hierarchical relationships among
  terms and concepts in a domain.
  (www.cordis.lu/ist/ka1/administrations/publication
  s/glossary.htm)

Definitions
Implementation
Metabolomics standards are just now being
formulated, and it is expected that it will be
years before they reach maturity. In the
meantime, it will be necessary to ensure that the
HMDB uses the latest standards, and supports the
migration of data between old and new
versions. In the same way, the ontologies that
the standards reference are dynamic, and the HMDB
must be able to adapt to changes in the
definitions and structures that make up
metabolomics shared vocabulary. This means that
our data architecture must be flexible,
extensible to new areas of research, and scalable
as the flow of data grows from a trickle to a
flood.

• CAS
• Beilstein
• KEGG

• NCBI
• PubChem
• MetaCyc