Disease Gene Finding.

1


Disease Gene Finding. Table of contents
Background Why do we want to find disease genes,
how has it been done until now? Networks
deducing functional relationships from network
theory Networks Biological networks Functiona
l modules / network clusters Phenotype
association Grouping disorders based on their
phenotype. Biological implications of phenotype
clusters. Method and examples Combining
network theory and phenotype associations in an
automated large scale disease gene finding
platform Proof of concept.
2


Abstract
Aim Find new disease genes. Means Use
protein interaction networks and phenotype
association networks for inferring phenotype
gneotype relationships. Proof Interesting
candidates are reported to experimental collaborat
ors who perform mutational analysis in patient
material.
3


Background
4


Background
Aim
Finding genes responsible for major genetic
disorders can lead to diagnostics, potential drug
targets, treatments and large amounts of
information about molecular cell biology in
general.
5


Background Methods for disease gene finding
post genome era (gt2001)
Mircodeletions
Translocations
Linkage analysis
http//www.rscbayarea.com/images/reciprocal_transl
ocation.gif
Fagerheim et al 1996.
http//www.med.cmu.ac.th/dept/pediatrics/06-intere
st-cases/ic-39/case39.html
6
(No Transcript)
7
(No Transcript)
8
Which is Keyser Söze ?
9


Background Bioinformatic methods for disease
gene finding post genome era (gt2001)
?
Grouping Tissues, Gene Ontology, Gene
Expression, MeSH terms .
(Perez-Iratxeta, Bork et al. 2002) (Freudenberg
and Propping 2002) (van Driel, Cuelenaere et al.
2005) (Hristovski, Peterlin et al. 2005)
10


Disease Gene Finding. Table of contents
Background Why do we want to find disease genes,
how has it been done until now? Networks
deducing functional relationships from network
theory Networks Biological networks Functiona
l modules / network clusters Phenotype
association Grouping disorders based on their
phenotype. Biological implications of phenotype
clusters. Method and examples Combining
network theory and phenotype associations in an
automated large scale disease gene finding
platform Proof of concept.
11


Networks and functional modules Deducing
functional relationships from network theory
12


Networks and functional modules Deducing
functional relationships from network theory
Network theory is boooooooooring
13


Networks
Text mining of full text corpora e.g PubMed
Central
http//www.biosolveit.de/ToPNet/screenshots/fig1.h
tml
14


Networks
Protein interaction networks of physical
interactions.
(Barabasi and Oltvai 2004).
15


Networks
Social Networks, The CBS interactome
(de Licthenberg et al.)
16


Genetically heterogeneous disorders and protein
interactions
(de Licthenberg et al.)
http//www.biosolveit.de/ToPNet/screenshots/fig1.h
tml
(Barabasi and Oltvai 2004).
(Barabasi and Oltvai 2004).
17


Genetically heterogeneous disorders and protein
interactions
(de Licthenberg et al.)
http//www.biosolveit.de/ToPNet/screenshots/fig1.h
tml
(Barabasi and Oltvai 2004).
(Barabasi and Oltvai 2004).
18


Genetically heterogeneous disorders and protein
interactions
Degree (k) Number of connections Protein
Number of interaction partners Social Number of
collaborators / friends
Degree distribution P(k) The probability that
a selected node has exactly k links Protein
probability of k interaction partners Social
Probability of k collaborators / friends
(Barabasi and Oltvai 2004).
19


Genetically heterogeneous disorders and protein
interactions
Clustering coefficient C(k) Average clustering
coefficient of all nodes with k links. The
average tendency of nodes to form clusters or
groups. Protein Tendency of interaction
partners to interact with each other Social
Tendency of collaborators / friends to be friends
/ collaborators of each other.
Hubs, connect distant parts of the network.
Ultra small world
(Barabasi and Oltvai 2004).
20


Genetically heterogeneous disorders and protein
interactions
Social Networks, The CBS interactome
(de Licthenberg et al.)
21


Genetically heterogeneous disorders and protein
interactions
Social Networks, The CBS interactome
(de Licthenberg et al.)
22


Genetically heterogeneous disorders and protein
interactions
23


Genetically heterogeneous disorders and protein
interactions
Network clustering
Functional modules
24


Genetically heterogeneous disorders and protein
interactions
Network clustering
Functional modules
Edge/physical interaction Node/protein
25


Genetically heterogeneous disorders and protein
interactions

• Grouping of proteins that are functionally
  undescribed. (30 of proteins in completely
  sequenced geneomes cannot be appointed to a
  specific biological function).
• 70-80 of interacting proteins share at least one
  function.
• Grouping of proteins based on function not
  biochemistry/sequence alignment.
• Correlation between mutation in interacting
  proteins and phenotype.
• Disease gene finding!!

Edge/physical interaction Node/protein
26


Disease Gene Finding. Table of contents
Background Why do we want to find disease genes,
how has it been done until now? Networks
deducing functional relationships from network
theory Networks Biological networks Functiona
l modules / network clusters Phenotype
association Grouping disorders based on their
phenotype. Biological implications of phenotype
clusters. Method and examples Combining
network theory and phenotype associations in an
automated large scale disease gene finding
platform Proof of concept.
27


Phenotype association
28


Phenotype association
Smith-Lemi-Opitz Syndrome
Birth weight lt2500gm Failure to thrive Short
stature Anteverted nares Bitemporal
narrowing Broad alveolar margins Broad, flat
nasal bridge Cataracts Cleft palate Dental
crowding Epicanthal folds Hypertelorism Hypoplasti
c tongue Large central front teeth Low-set
ears Microcephaly Micrognathia Posteriorly
rotated ears Ptosis Strabismus Autosomal
recessive Elevated 7-dehydrocholesterol
Hypoplastic lungs Incomplete lobulation of the
lungs Hip dislocation Hip subluxation Limb
shortening Metatarsus adductus Overriding
toes Postaxial polydactyly Proximally placed
thumbs Short thumbs Short, broad toes Stippled
epiphyses Syndactyly of second and third
toes Talipes calcaneovalgus Blonde
hair Eczema Facial capillary hemangioma Severe
photosensitivity Shrill screaming
Constipation Malrotation Poor suck Pyloric
stenosis Vomiting Atrial septal
defect Coarctation of aorta Patent ductus
arteriosus Ventricular septal defect Ambiguous
genitalia Bifid scrotum Cryptorchidism Cystic
kidneys Hydronephrosis Hypoplastic
scrotum Hypospadias Micropenis Microurethra Renal
agenesis Single kidney Ureteropelvic junction
obstruction
Low cholesterol Allelic with Rutledge lethal
multiple congenital anomaly syndrome Estimated
incidence 1/20,000 - 1/40,000 Caused by mutations
in the delta-7-dehydrocholesterol reductase
gene Abnormal sleep pattern Aggressive
behavior Frontal lobe hypoplasia Hydrocephalus Hyp
ertonia (childhood) Hypotonia (early
infancy) Mental retardation Periventricular gray
matter heterotopias Seizures Self injurious
behavior Breech presentation Decreased fetal
movement
29


Phenotype association
Word vectors
(Brunner and van Driel 2004)
30


Phenotype association
Word vectors
(Brunner and van Driel 2004)
31


Phenotype association
Word vectors
32


Phenotype association
Word vectors
33
(No Transcript)
34


Phenotype association
Word vectors
ACHOO SYNDROME 100820
Gastric Sneezing 137130 0.441407
(Brunner and van Driel 2004)
35
(No Transcript)
36


Disease Gene Finding. Table of contents
Background Why do we want to find disease genes,
how has it been done until now? Networks
deducing functional relationships from network
theory Networks Biological networks Functiona
l modules / network clusters Phenotype
association Grouping disorders based on their
phenotype. Biological implications of phenotype
clusters. Method and examples Combining
network theory and phenotype associations in an
automated large scale disease gene finding
platform Proof of concept.
37


Method Proof of concept
38


Method
39


Method
40


Proof of Concept
Input all critical intervals in OMIM (Approx 900)
125480 MAJOR AFFECTIVE DISORDER 1 132800
MULTIPLE SELF-HEALING SQUAMOUS EPITHELIOMA 137100
IMMUNOGLOBULIN A DEFICIENCY SUSCEPTIBILITY
1 137580 GILLES DE LA TOURETTE SYNDROME 143850
ORTHOSTATIC HYPOTENSIVE DISORDER 156240
MESOTHELIOMA, MALIGNANT 157900 MOEBIUS SYNDROME
1 177900 PSORIASIS SUSCEPTIBILITY 1 209850
AUTISM 252350 MOYAMOYA DISEASE 1 608631
ASPERGER SYNDROME, SUSCEPTIBILITY TO, 2
ASPG2 301845 BAZEX SYNDROME BZX 608389
BRANCHIOOTIC SYNDROME 3 600175 SPINAL MUSCULAR
ATROPHY 600318 DIABETES MELLITUS,
INSULIN-DEPENDENT, 3 IDDM3 INSULIN-DEPENDENT
DIABETES MELLITUS 3 601042 CHOREOATHETOSIS/SPASTI
CITY 601388 DIABETES MELLITUS,
INSULIN-DEPENDENT, 12 IDDM12 INSULIN-DEPENDENT
DIABETES MELLITUS 12 601493 CARDIOMYOPATHY,
DILATED, 1C CMD1C 603694 DIABETES MELLITUS,
NONINSULIN-DEPENDENT, 3 NIDDM3
NONINSULIN-DEPENDENT DIABETES US 3 604288
CARDIOMYOPATHY, DILATED, 1H CMD1H
41


Proof of Concept
Input all critical intervals in OMIM (Approx 900)
125480 MAJOR AFFECTIVE DISORDER 1 132800
MULTIPLE SELF-HEALING SQUAMOUS EPITHELIOMA 137100
IMMUNOGLOBULIN A DEFICIENCY SUSCEPTIBILITY
1 137580 GILLES DE LA TOURETTE SYNDROME 143850
ORTHOSTATIC HYPOTENSIVE DISORDER 156240
MESOTHELIOMA, MALIGNANT 157900 MOEBIUS SYNDROME
1 177900 PSORIASIS SUSCEPTIBILITY 1 209850
AUTISM 252350 MOYAMOYA DISEASE 1 608631
ASPERGER SYNDROME, SUSCEPTIBILITY TO, 2
ASPG2 301845 BAZEX SYNDROME BZX 608389
BRANCHIOOTIC SYNDROME 3 14q23.1
SIX1 600175 SPINAL MUSCULAR ATROPHY 600318
DIABETES MELLITUS, INSULIN-DEPENDENT, 3 IDDM3
INSULIN-DEPENDENT DIABETES MELLITUS 3 601042
CHOREOATHETOSIS/SPASTICITY 601388 DIABETES
MELLITUS, INSULIN-DEPENDENT, 12 IDDM12
INSULIN-DEPENDENT DIABETES MELLITUS 12 601493
CARDIOMYOPATHY, DILATED, 1C CMD1C 10q21-q23
VINC_HUMAN 603694 DIABETES MELLITUS,
NONINSULIN-DEPENDENT, 3 NIDDM3
NONINSULIN-DEPENDENT DIABETES US 3 604288
CARDIOMYOPATHY, DILATED, 1H CMD1H
42


608389 BRANCHIOOTIC SYNDROME 3 14q23.1
SIX1
43


Proof of Concept
  SIX1 mutations cause branchio-oto-renal
syndrome by disruption of EYA1-SIX1-DNA
complexes.Ruf RG, Xu PX, Silvius D, Otto EA,
Beekmann F, Muerb UT, Kumar S, Neuhaus TJ, Kemper
MJ, Raymond RM Jr, Brophy PD, Berkman J, Gattas
M, Hyland V, Ruf EM, Schwartz C, Chang EH, Smith
RJ, Stratakis CA, Weil D, Petit C, Hildebrandt
F.Department of Pediatrics, University of
Michigan, Ann Arbor, MI 48109, USA.Urinary
tract malformations constitute the most frequent
cause of chronic renal failure in the first two
decades of life. Branchio-otic (BO) syndrome is
an autosomal dominant developmental disorder
characterized by hearing loss. In
branchio-oto-renal (BOR) syndrome, malformations
of the kidney or urinary tract are associated.
Haploinsufficiency for the human gene EYA1, a
homologue of the Drosophila gene eyes absent
(eya), causes BOR and BO syndromes. We recently
mapped a locus for BOR/BO syndrome (BOS3) to
human chromosome 14q23.1. Within the 33-megabase
critical genetic interval, we located the SIX1,
SIX4, and SIX6 genes, which act within a genetic
network of EYA and PAX genes to regulate
organogenesis. These genes, therefore,
represented excellent candidate genes for BOS3.
By direct sequencing of exons, we identified
three different SIX1 mutations in four BOR/BO
kindreds, thus identifying SIX1 as a gene causing
BOR and BO syndromes. To elucidate how these
mutations cause disease, we analyzed the
functional role of these SIX1 mutations with
respect to protein-protein and protein-DNA
interactions. We demonstrate that all three
mutations are crucial for Eya1-Six1 interaction,
and the two mutations within the homeodomain
region are essential for specific Six1-DNA
binding. Identification of SIX1 mutations as
causing BOR/BO offers insights into the molecular
basis of otic and renal developmental diseases in
humans.PMID 15141091 PubMed - indexed for
MEDLINE
44


604288 CARDIOMYOPATHY, DILATED, 1C CMD1C
10q21-q23 VINC_HUMAN
45


Proof of Concept
Metavinculin mutations alter actin interaction
in dilated cardiomyopathy. Olson TM, Illenberger
S, Kishimoto NY, Huttelmaier S, Keating MT,
Jockusch BM. Department of Pediatrics and the
Division of Cardiology, University of Utah, Salt
Lake City, Utah, USA. olson.timothy_at_mayo.edu BACK
GROUND Vinculin and its isoform metavinculin are
protein components of intercalated discs,
structures that anchor thin filaments and
transmit contractile force between cardiac
myocytes. We tested the hypothesis that heritable
dysfunction of metavinculin may contribute to the
pathogenesis of dilated cardiomyopathy (DCM).
METHODS AND RESULTS We performed mutational
analyses of the metavinculin-specific exon of
vinculin in 350 unrelated patients with DCM. One
missense mutation (Arg975Trp) and one 3-bp
deletion (Leu954del) were identified. These
mutations involved conserved amino acids, were
absent in 500 control individuals, and
significantly altered metavinculin-mediated
cross-linking of actin filaments in an in vitro
assay. Ultrastructural examination was performed
in one patient (Arg975Trp), revealing grossly
abnormal intercalated discs. A potential
risk-conferring polymorphism (Ala934Val),
identified in one DCM patient and one control
individual, had a less pronounced effect on actin
filament cross-linking. CONCLUSIONS These data
provide genetic and functional evidence for
vinculin as a DCM gene and suggest that
metavinculin plays a critical role in cardiac
structure and function. Disruption of force
transmission at the thin filament-intercalated
disc interface is the likely mechanism by which
mutations in metavinculin may lead to DCM.
46
How well does it work ?
47
How well does the score work ?
48
Is it unbiased ?
49
Reveals novel global aspect of human diseases
50


Disease Gene Finding. Summery
Background Why do we want to find disease genes,
how has it been done until now? Networks
deducing functional relationships from network
theory Networks Biological networks Functiona
l modules / network clusters Phenotype
association Grouping disorders based on their
phenotype. Biological implications of phenotype
clusters. Method and examples Combining
network theory and phenotype associations in an
automated large scale disease gene finding
platform Proof of concept.
51


Hopes Dreams for the future
52


Acknowledgements
Disease Gene Finding group at CBS Olga Rigina
Database handling, Computer Scientist Olof
Karlberg Programmer, Pharmacologist Zenia M.
Larsen Expert in diabetes and related
disorders, Engineer Páll Ísólfur Ólason
Engineer, data flow, text mining. Kasper Lage
Proteomics, genomics, diseases, Human
Biologist Anders Hinsby Proteomics, mass
spec. expert, Human Biologist