Linkage non random assortment of genes at meiosis, usually detected as nonrandom segregation in offs

1

• Linkage - non random assortment of genes at
  meiosis, usually detected as non-random
  segregation in offspring.
• Why is it useful?
• To find the cause of diseases showing Mendelian
  or near-Mendelian inheritance.
• To demonstrate genetic heterogeneity
• To predict risk of disease when specific mutation
  not known.(still useful in DMD and CF).
• To exclude risk of disease
• Linkage studies are getting easier!

2
Possible gametes
50 Recombination - random number of cross-overs
between genes. Very far apart or on different
chromosomes.
3
Recombination greater than 50 not seen,since
second crossover may involve any strand In
practice VERY large samples required to detect
recombination fractions of gt 0.25 as different
from random. Very large samples needed to
measure recombination at lt 1 ? 1 megabase
(very approx.) Genetic distance measured in
centimorgans.1cM is a distance in which there is
a 1 chance of a cross over in a given
strand. For short distances eg lt 5 recombination
,RF and cM numerically the same (double
recombinants VERY rare in 5cM)
4
(No Transcript)
5

• Sources of information for linkage
• .3 generation families
• .2 generation families
• .Sib-pairs - sharing of haplotypes e.g. in
  diabetes (mostly for complex traits)
• .Single individuals
• Typing 2 loci on single sperm
• Homozygosity mapping

Populations of unrelated individuals show
ancestral record of linkage through linkage
disequilibium
6

• Consider detection of linkage of an autosomal
  dominant disease to a co-dominant marker with
  all members of family tested.
• Requirement Genetic variation - parent must be
  double heterozygote.
• Assume homogeneity i.e. disease is caused by
  mutation in same gene in different families,
  although not necessarily same mutation

7
A phase-known family
If ? 0 likelihood of this result is (1- ?)4
(1)4 1 If ? 0.5 likelihood of this result is
(0.5)4 1, i.e. 1/16
8
Therefore, likelihood of this result if ? 0
likelihood of this result if ? 0.5 lod i.e.
log of odds log16 1.20 If recombination
fraction is 10 i.e. ? 0.1, likelihood of this
family (? -1)4 (0.9)4 0.65 Therefore,
likelihood of this result if ? 0.1
likelihood of this result if ? 0.5 lod i.e.
lod of odds log10.49 1.02 lod scores at
different values of ? 0 0.1 0.2 0.3 0.4 0.5
Family 1 4NRCS 1.20 1.02 0
9
Family 2. Phase-known
If recombination fraction is 10 i.e. ? 0.1,
likelihood of this family (1-? )nr X (?)rc
(0.9)4 X (0.1)1 0.065 lod log (0.065 /(1/ 32
)) lod scores at different
recombination fractions 0 0.1 0.2 0.3 0.4 0.5 Fa
mily 2. 4NR 1R -? 0.32 0.42 0.36 0.22 0
10
Why lod scores? If all families like this. ?no
need for complex statistics. Just count
recombinants and add from different
families. Some families like this CEPH - very
good for markers. Mice also have big families and
dont need lodscores But most disease families
not like this at all!
11
Family 3. Phase not known
12
Add results for three families Lod scores at
? 0 0.1 0.2 0.3 0.4 0.5 Fam 3. Z1 30
0.602 0.465 0.318 0.170 0.049 0 Fam 2.
4NR 1 RC -? 0.321 0.418 0.362 0.219
0 Fam 1 4 NR 1.204 1.020 0.816
0.584 0.316 0 -? 1.806
1.552 1.116 0.584 0 Families are independent
so probabilities can be multiplied, i.e. lod
scores can be added. Ability to add lod scores
from different sources and types of families is
main advantage of lods.
13
Some Families to Score
3NRCS
1 NRC
14
Two more families
No score
Z1 31
15
(No Transcript)
16
Unless family is CONSANGUINOUS -gt HOMOZYGOSITY
MAPPING.
17
Interpretation of a lod-score A lod-score of 3
at a given value of ? means that the results
obtained are 1000 times more likely if the loci
are linked at the given ? than if they are not
linked. lod gt 3 at any ? taken to mean loci are
linked (about 2 false positives). lod lt- 2 at
any ? - loci not linked at that ?. lod score
between - need more data. In theory 10
phase-known meioses can give a lod score of 3 at
? 0 (but in practice more are usually needed).

18
Height of lodscore reflects confidence that
linkage exists. The highest lodscore does not
always indicate that this is the nearest
marker,it may just be the most informative. If
you have a prior hypothesis a lodscore of 2 is
very exciting! A very small family can disprove
linkage
19

• Lod scores are easiest to use when
• .Penetrance is high
• No phenocopies
• .Disease is genetically homogeneous
• .Disease is rare
• .Disease does not affect reproduction
• .Disease has onset in early life
• The ideal situation is one large family!

20
Important points about accepting a lod score of
3 as evidence for linkage Hypothesis about mode
of inheritance, penetrance and phenocopy rate
made BEFORE analysing data. Diagnosis (affected,
unaffected or unknown) made independently
preferably BEFORE analysis. Families or
individuals may be excluded because of
uncertainty about any ONE finding (ie either
phenotype or genotype but not because you dont
like the combination) Affecteds-only analysis is
not biassed but loses power. The most trouble in
linkage analysisis caused by ?
WRONG DIAGNOSIS
21
Points to Ponder .Why is false rate 2, why
not 1 in 1000? .Why is the value of the
lod-score at ? 0 not always the most
useful? .Why use a computer? What comes next ?
Pinpointing the gene
22
A family wit h Tuberous sclerosis thought to have
non-penetrance
23
A problem from Guys Genetics centre
Rachel King has undergone prenatal diagnosis in
order to rule out the possibility of Huntington
disease in her 11 week old fetus. Rachel is
currently unaffected by Huntington disease and
does not wish to find out whether she is likely
to develop symptoms herself or not. Her father,
Mr Gray, died of Huntington disease several years
ago but no sample is available from him. However,
DNA samples have been tested from Rachels mother
(Mrs Gray), Rachels husband (Daniel King),
Rachel, and her pregnancy (a chorionic villus
sample). Two dinucleotide repeat markers have
been used, D4S126 and D4S127, which are very
closely linked to the Huntington disease locus on
chromosome 4. The allele sizes obtained are given
in the table below. Marker results (Alleles are
shown as sizes in base pairs) D4S126 D4S127 R
achel King 194, 200 147, 149 Daniel
King 196, 198 149, 151 Mrs Gray 194,
198 145, 147 CVS 196, 200 151, 149
What is the risk that this pregnancy will be
affected by Huntington disease ? What is the risk
that Rachel King will develop Huntington disease
later in life ?
24
Slide from Guys Genetic centre