Relationship Testing Flashcards
What can relationship testing also be known as?
Kinship analysis
What is relationship testing used for?
- Used to assess whether (and how) people may be genetically related to one another
- Relatedness calculations are carried out based on number of alleles shared
What are some applications for relationship testing?
o Paternity testing
o Immigration cases
o Missing person identification (disaster victim identification)
o Inheritance disputes
In relationship testing, what is it that is being investigated?
Genetic relatedness
What evidence are we searching for with relationship testing?
- We are looking for evidence to suggest that one individual is direction descended from the other
o Children are separated from their parents by **one generation **
o Grandchildren and separated from their grandparents by **two generations ** - Or that individuals share a very **recent common ancestor **
o Full siblings share two parents
o Half-siblings share one parent
o First cousins share two grandparents
Whats recombination?
How was relationship testing first performed?
- First performed **using blood groups **
o Very useful for excluding a man as the potential father as mutations do not occur
Why are mutations more common in STR than in blood groups?
STR in bit of DNA non-coding and blood groups are proteins, there for a purpose, for survival and function
What was the first application of ‘DNA fingerprinting’?
**First application of ‘DNA fingerprinting’ (RFLP testing) was an immigration case (maternity) **
o A boy born in Britain to a Ghanian mother had just returned to the UK after living with his father in Ghana for a short while
o About 20 blood groups- one exclusion (Duffy)
o Mother Christiana phenotype Fy(a+)
o Child Andrew phenotype Fy(b+)
- Certainly, closely related, but was Christiana his aunt?
o Bands shared on average 26% of the time - Mother v aunt more than one billion times more likely
- Currently routinely carried out using STR typing
In what situations would you want to determine paternity (maternity)?
Give subgroup examples for each
What are the non-paternity rates?
- Quoted rates vary 1%-30%
o Methodology often not validated
o Prevalent in testing laboratories- 15%
Paternity questioned
o Recent studies
**1-3% estimated in the UK ** - Large variation seen between countries, cultures and different social statis
How is the assessment of paternity done in the following case?
- Mother names a man as the father of her child
- If the ‘mother’ is the true mother, then mother and child must have an allele in common
- Tested man is not excluded from paternity
What is the equation for likelihood ratio?
What is the chance of the tested man being the father?
How would you work out the probability of independent events?
- When both A and B occur together
A n B= p(A) x p(B)
What do indepent events allow in forensic genetic situations?
What are some limitations to this?
- Allows likelihood ratios from independent genetic systems to be combined into a total LR by multiplication
o 1.89 x 1.45 x ….. = 10,0000 (minimum)
o The **‘combined paternity index (CPI)’ **
*** Caveats **
o Same chromosome- genes may be linked (VWA and D12S391 are linked)
o Different chromosomes- non-random assortment due to population structure
Tell me about the paternity index (likelihood of paternity)
**Paternity index (likelihood of paternity)= 10,000 **
* The tested man is 10,000 times more likely to be the father than an unrelated man (no)
- The types observed are10,000 time more likely to be seen in a trio of true father, mother, and child than in a trio of an unrelated man, mother and child
- A laboratory should define a minimum PI- 100 standard in US- most labs look for at least 10,000
What does the report day regarding PI?
**Some paternity reports may say: **
o Paternity index (LR)= 10,000
LR= likelihood ratio
o Probability of paternity (W)= 99.99%
W (Wahrscheinlichkeit)- probability
o Implies that these are two ways of expressing the same thing, which they are NOT
Given the evidence, what are the offs that the man we have tested IS the father?
How is this caluclated and what is the equation for this?
Tell me what Prior odds and what they depend on
A range of priors?
- Assuming that the CPI is 10,000
- ‘Assuming a prior of between 10% and 90% the paternity index is somewhere between 1,111 and 900,000’
o 10% prior= 0.1/0.9 x 10,000 = 1,111
o 90% prior= 0.9/0.1 x 10,000= 90,000 - A pragmatic approach
o We say- 10,000, ‘assuming no other relevant evidence’
But is that 99.99%?
- Probability of paternity (W)
= **(Pi LR) / (Pi LR + 1) – (Essen-Möller equation) **
Where Pi is the prior odds
Assuming a 50% prior
o **Probability (W) = LR / LR+1 **
o If paternity index = 10,000
Probability of paternity= 10,000 / 10,001
= 0.9999
= 99.99%
o Can never reach 100% and always include non-genetic evidence- unlike the PI
What are the pros/cons of paternity of a diagnositc test?
- A good diagnostic test is one which will minimise the number of false positive and false negative results
- **False positive- **telling someone that they have a disease, when in fact they do not
o Significant false positive rates for mammograms
* False positive- telling someone that they are the father of a child when they are not
o Not tolerable- an easy mistake in some circumstances
* False negatives- telling someone that they do not have the disease, when in fact they do
o Significant false negative rates for cervical cytology
o Significant false negative rates for COVID19
* False negatives- telling someone that they are not the father of a child, when they are
o Not tolerable- but an easy mistake
* Unlike diagnostic tests- the ‘one off’
False positive paternity
- Man, mother, and child tested
- CPI= 21,000
- Probability of paternity= 99.995%
- PI is the genetic odds in favour of paternity
- Man is the brother of the true father
- “Offers very strong evidence paternity providing a close relative of the tested man is not a possible father”
