• Congratulations to the Class of 2024 on your results!
    Let us know how you went here
    Got a question about your uni preferences? Ask us here

why is dna fingerprinting more accurate than blood types for paternity testing? (1 Viewer)

Eagle Mum

Well-Known Member
Joined
Nov 9, 2020
Messages
561
Gender
Female
HSC
N/A
Test accuracy comprises sensitivity & specificity which affect positive and negative predictive values of results.

It’s often quite easy to prove non paternity. A male with blood type O cannot have fathered a child with AB blood type unless the mother of the child has AB type as well, but it is very hard to ‘prove’ paternity satisfactorily for legal purposes based on only the eight possible combinations (A+, A-, B+, B-, AB+, AB-, O+, O-) provided by the standard blood group typing system, with four possible options for the A/B antigens and two options with the Rhesus system. Even if there’s a match with the rarest blood type (AB-), at 0.6% prevalence, there‘d be approx. 1 in 166 chance of a match between unrelated individuals which would still be thousands of potential paternal candidates in the population.

On the other hand, the lengths of short tandem repeat sequences usually vary widely amongst individuals, and half a dozen microsatellite loci, each with a range of ten, twenty or more repeating units, can be used to confirm paternity up to a probability of 1 in 20 million of non paternity. For example, if at a single locus, individuals in a population each carry 10-17 repeating units on each allele, then there is an average probability of 1 in 8 that two unrelated individuals would match for the paternal allele (noting that some lengths occur more frequently than others, so there might actually be 5-20% chance of a match between unrelated individuals). With seven or eight such microsatellite foci included in sample analysis, paternity could potentially be established at a probability of 1 in a million or greater that the match occurred randomly between unrelated individuals, by multiplying all the probabilities for random matches at each STR, so a perfect match is usually accepted as evidence unless the candidate has an identical twin.

In short, the blood group system offers three binary options (ie. an individual either has or doesn’t have an A, B and/or RhD antigen), whilst you can select micro/minisatellites, so that there are many possible lengths for each STR.
 
Last edited:

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Top