DNA paternity testing – how it works and the risks involved
I recently read a detailed article in New Scientist about some labs in the USA faking DNA paternity tests on unborn children and the repercussions of this on people’s lives. These labs were using very questionable science to assess fatherhood, often leading to false results and sometimes confusing findings.
The reason they had become popular was the ease with which you could get this test. Whereas a normal test of a child in the womb would require an amniocentesis (a considerably invasive procedure which involves injecting a long needle into the mother’s stomach to access the amniotic fluid around the foetus), all these labs required was a sample of the mother’s blood and swabs from all potential fathers’ cheeks.
After hearing some disturbing cases, New Scientist investigated more closely, sending in their own samples. In a particularly strange case, a sample was sent in from a non-pregnant woman with swabs from two random men’s’ checks. One of these labs sent back a result containing the non-existent foetus’s genetic information.
This got me thinking, how exactly do paternity tests work and how were these labs able to fake their results? The process of identifying genetic relationships between individuals is called DNA fingerprinting. This is why. Where fingerprints are used as a small … to distinguish individuals, this process only uses a fragment of DNA to identify links between people.
Whereas sequencing whole genomes is a lengthy and costly process, it is possible to select specific areas of DNA to analyse. The most common procedure is to analyse areas known as STRs (Short Tandem Repeats), which are, repeats of nucleotide sequences ranging from 2-50 nucleotides. Here’s an example. In the following DNA sequence, the pattern of nucleotides CATG is repeated 4 times: GTCAGTCATCATGCATGCATGCATGTCAGC.
By identifying these specific repeats in a specific region of the genome, a genetic profile can be created for the individual in question. At the present time, over 10,000 STRs have been published in the human genome. By comparing a selection of these regions between individuals, it is possible to detect whether they are related. The reason these regions are of particular use is that they are often found in regions of DNA that does not code for anything (better known as introns, they are found in between the exons that code for specific genes).
They are therefore less likely to undergo evolution and change within a species, making them a perfect marker of the human genome. In forensics, sequences of no more than 4 to 5 nucleotides are more often used. Shorter sequences are more likely to change during the process used to analyse them and longer sequences could suffer from degradation. In paternity testing a male child, comparing sequences from their Y chromosome is of particular interest and it is passed directly from father to son.
To create a genetic profile for an individual, DNA will be extracted from the cells of a sample. The sequences of interest will be amplified using a process called PCR (Polymerase Chain Reaction), then analysed through means of a gel to determine how many repeats of the STR sequence there are. These profiles are stored in databases and different countries have different rules regarding what they consider a proper analysis. The US for example demands the analysis of 13 specific STR sequences (or loci); the UK uses 10.
So how is it possible to fake these results? With poor science is the answer. Many of the labs investigated by New Scientist failed in many aspects of rigour and technique. One lab stopped using STRs and opted for SNPs (Single Nucleotide Polymorphisms), which are not a good tool for this analysis due to the repeat of a single nucleotide being a lot more likely than that of a longer sequence.
Investigation into another team’s results showed their analysis of the Y chromosome STRs to be completely off. As a chromosome that has no partner, the Y chromosome is easier to analyse and its ancestry and evolution has been mapped. Y-chromosomes are set out on evolutionary branches, yet the labs results placed one of the individual’s genetic data on several branches, which is simply put impossible. See the New Scientist article – The dangers of unreliable paternity testing – for more details.
Much controversy is still surround genetic fingerprinting, whether it is used to solve crime, assess paternity or analyse fossils and bone. A recent controversy arose around the analysis of DNA from the fossilised bones of Tutankhamen. Scientists have unveiled their analysis of the king’s lineage but others are sceptical of the methods used, claiming more research needs to be done and that the details of the data analysed should be released.
Although genetic fingerprinting has come far, it still needs to specify more clearly what defines an individual’s genetic profile and make sure the proper methods are being used by all labs involved, especially organisations making profit.