Transcript Respective contributions of MIAME, GeneOntology and UMLS

Respective contributions of MIAME,
GeneOntology and UMLS for
transcriptome analysis
Fouzia Moussouni, Anita Burgun, Franck Le Duff,
Emilie Guérin, Olivier Loréal
INSERM U522 and Medical Informatics Laboratory,
CHU Pontchaillou
Rennes, FRANCE
Transcriptome & DNA microarray
study of transcriptionnal response of the cell
Normal
Pathologic
Response to
chemics or foods
treatment
Response to a
growth factor
Response to
genetic
disturbances
Pathological situations studied at INSERM U522
 DNA mutation(s)
Hemochromatosis…
 Chronic liver diseases
IRON
overload
Mechanisms
 Fibrosis
 Cirrhosis
 Hepatocarcinoma
One may deposit thousands of genes
1 measure
1 Expression Level
1 Spot intensity
Intensive data generation
1 gene
but multiple
facets !
Experimental
Raw Data
Available knowledge on the
expressed genes, that need to be
capturized and organized.
One gene but multiple descriptions
 Nucleic Sequence components - promoters, introns, exons, transcripts, regulators, …
 Chromosomal localization,
 Functional proteins and known genes products,
 Tissue distribution,
 Known gene interactions,
 Expression level in physiologic and pathologic conditions,
 Known gene variations,
 Clinical Implications,
 Literature and bibliographic data on a gene.
Need of an integrated gene expression
environment (for the liver!)
External
Sources
? ?
?
Data cleaning !
Clinical
Data
Integration
experimental
data
SAGE
Micro-arrays Substractive
banks
Gene
Expression
warehouse
Analysis
BIO
KNOWLEDGE
Gene Expression
Warehouse
Standardization
and controlled
specification
ONTOLOGY DESIGN
Knowledge extraction and data exchange
Standardization
ONTOLOGY DESIGN
Respective contributions
MIAME
GO
UMLS
MIAME
MIAME will provide a standard framework to represent
the minimum information that must be reported about
microarray experiments :
•
•
•
•
•
•
Experience
Work in progress ...
Array
Samples
Hybridization
Measures
Normalisation and control
Minimum information about a microarray experiment (MIAME) toward standards for
microarray data', A. Brazma, at al., Nature Genetics, vol 29 (December 2001), pp 365 371.
GeneOntology
(GO)
GO is an ontology for molecular biology and
Genomics,
But GO is not populated with :
 gene sequences
 gene products, ...
GOA
UMLS
 The Unified Medical Language System
(UMLS) is intended to help health professionals
and researchers to use biomedical information
from different sources.

Examples from iron metabolism are
studied
 How pathologic disease states related to
iron metabolism alteration are described in
GO and UMLS ?
BIOLOGICAL MODEL FOR IRON METABOLISM
IRON
METABOLISM
GENES
alteration
PATHOLOGIC
STATES
Iron metabolism
diseases
Iron overload
aceruloplasminemia
Iron deficiency
Other diseases
hyperferritinemia
cataract
Iron overload due to a gene alteration
Iron overload during Aceruloplasminemia
Gene
Ceruloplasmin
mutation
NO
NO
Feroxydase activity in plasma
Fe2+
Fe3+
Iron binding with plasmatic transferrin
THE IRON STAYS INSIDE THE CELL !!
BIOLOGICAL MODEL FOR IRON METABOLISM
IRON
METABOLISM
GENES
alteration
PATHOLOGIC
STATES
Iron metabolism
diseases
Iron overload
aceruloplasminemia
Iron deficiency
Other diseases
hyperferritinemia
cataract
A second scenario related to iron metabolism genes alteration
Cataract and hyperferritinemia
mRNA
gene
L_Ferritin
L_Ferritin
mutation
IRE
Translation
in excess
IRP
L_Ferritin protein
in excess
CATARACT and
HYPERFERRITINEMIA !
UMLS view
Cataract and hyperferritinemia
Iron compound
AA, Peptide or Prorein
Biologically Active Substance
Metalloprotein
Ferritin
AA, Peptide or Protein
H_Ferritin
L_Ferritin
Co-occurs
In Medline
IRE
RNAbinding
Protein
Iron
Sulfur
Prot
Co-occurs
In Medline
(freq 26)
Cataract
IRP
GO/ GOAnnotations view
Cataract and hyperferritinemia
Cell component
Ligand binding Prot or carrier
Ferritin
Ferric iron binding
Link in GO Annotations DB
Ferritin
Heavy
Chain
Iron homeostasis
Iron transport
Ferritin Light Chain
Metabolism
IRE
IRP
Hydro-lyase
Cataract
Target representation
Cataract and hyperferritinemia
Ligand binding Prot or carrier
Ferritin
Ferric iron binding
Iron homeostasis
Ferritin
Heavy
Chain
Iron transport
Ferritin Light Chain
Genes
Mutated genes
IRE
IRP
Hyperferritinemia
Dynamic links
Modeling of biological functions
Cataract
And more generally …
Recapitulative
Information on disease states,
clinical treatments and
followups.
Normal vs. pathologic
UMLS
DNA Chips
Information M
on biological I
samples,
A
experiments
and results M
E
We need precise and
dynamic models to get
the whole picture
?
GOA
Information on Roles of the
genes in Biological and
metabolic states
Gene products for Iron metabolism, as they
are actually described in GO and UMLS.