Transcript Medical Genetics 1

Medical Genetics 1
Prof Duncan Shaw
http://www.abdn.ac.uk/~gen155/djshome.html
Major Groups of Clinical Disorders
with a Genetic Contribution
•
•
•
•
Single gene defects
Chromosomal abnormalities
Congenital malformations
Multifactorial diseases - most common
causes of illness
Autosomal recessive inheritance
• Cystic fibrosis
(1/2000)
• Recessive mental
retardation
(1/2000)
• Congenital
deafness (1/5000)
Increased risk in autosomal
recessive disease
• Consanguinity: if parents are related
(consanguinity) there is an increased risk that
both parents carry the same recessive allele
Degree Example
% alleles shared
1st
Siblings, parent/child
50
2nd
Uncle/niece,
grandparent/child
1st cousins
25
3rd
12.5
Ethnic associations with AR
disease
• In particular populations, recessive allele
frequency may have increased by selection in
heterozygotes, or by genetic drift
• -Thalassaemia: Cypriots, Greeks, Italians,
Chinese, African-Americans
• Sickle Cell Disease: Arabs, West Indians
• Tay-Sachs Disease: Ashkenazi Jews (4%
carriers)
• Severe Combined Immunodeficiency Syndrome:
Apache Native Americans
• Cystic Fibrosis: Caucasians
Finding the cystic fibrosis gene
• CF gene was found using positional
cloning
• Linkage to markers on chromosome 7
• But that didn’t get closer than several Mb –
still lots of genes
• To narrow the candidate region further,
used linkage disequilibrium…..
Linkage and linkage disequilibrium
• Linkage is tested within families, LD by population study
• This marker is linked to the disease, but to different
alleles (of the same marker gene) in each family
1,2
1,1
1,2
1,1
2,2
1,1
1,2
2,2
1,2
2,2
1,1
2,2
How LD
arises
LD and haplotypes
• Haplotype – the set of alleles carried by an
individual chromosome
• With N bi-allelic markers, expect 2N possible
haplotypes in population, because
recombination creates all possible combinations
of alleles
• If fewer than 2N haplotypes are observed, this is
evidence for LD
• Previous example: A1/A2 and CF/N gives 4
haplotypes with recombination, or 3 with LD
Testing for LD
Patients
Marker A (shows
LD with disease)
Marker B (no LD,
but could be linked)
Allele 1
Allele 2
Allele 1
Allele 2
150
50
120
80
100
120
80
Controls 100
c2 test for significance
LD operates over short genetic
distances
1
LD
0
-5000
-100
0 +100
Distance (kb) from disease gene
+5000
Use of LD for gene mapping
• A gene can be mapped by linkage in families to
within a few cM ( = a few Mb in humans)
• If all or most cases of the disease are
descended from a unique mutation, LD will be
observed with markers about 100kb or less from
the gene – much closer than you can get using
linkage alone
• In CF, about 70% of mutations are the same
(DF508) and these show LD with markers very
close to the CF gene – this helped the gene to
be identified
Autosomal dominant inheritance
• An affected
person usually
has one affected
parent
• Transmitted by
either sex
• Child of an
affected parent is
at 50% risk of also
being affected
Autosomal Dominant Diseases
Disease:
Frequency/1000 births:
Otosclerosis
Familial hypercholesterolaemia
Adult polycystic kidney disease
Multiple exostoses
Huntington’s disease
3
2
1
0.5
0.4
Multiple exostoses
The ear
Comparisons between AD and AR
Dominant
Recessive
• Expressed in
heterozygote
• Approx. 1/2 offspring
affected
• Equal frequency and
severity in each sex
• Paternal age effect on
rate of new mutation
• Variable expressivity
• Expressed in homozygote
• Low risk to offspring
• Equal frequency and
severity in each sex
• New mutations rare
• Constant expressivity in
each family
• Importance of
consanguinity
Revision of linkage and Lod scores
• Affecteds have A marker allele from Dad,
unaffecteds have B
• If random, would expect 50:50 distribution
• Evidence for linkage?
Revision of linkage and Lod scores (2)
• If marker and disease were unlinked, probability of this
pedigree: (1/2)4 = 1/16 = 0.0625
• If they are linked with RF = 0.1 (10% recombination),
probability of pedigree: (0.9)4 = 0.66 and odds ratio
(relative to no linkage) = 0.66/0.0625 = 10.56
• If they are linked with RF = 0.0, probability of pedigree:
(1)4 = 1 and odds ratio (relative to no linkage) =
1/0.0625 = 16
• To combine information from several families, take log10
of odds ( = LOD score) and add them up
• LOD > 3 good evidence for linkage; LOD < -2 evidence
against linkage; -2 < LOD < 3 is inconclusive