's the 'fingerprint' unique? Doubt over DNA
Socialist Organiser1992 1992-01-16Read it in the scan (PDF), scan page 14 →
Read from the scan by machine and not corrected: expect misread words, and now and then a caption or a line from a neighbouring story. For anything you quote, read the scan.
Scan page 14
Les Hearn's
even years ago, a powerful new forensic test was introduced - DNA "fingerprinting". Since then many criminals have been convicted and innocent people exonerated on the basis of DNA testing.
However, small amounts of doubt have crept in about its reliability in a few cases.
The test does not involve analysing the whole of the DNA. That would take several years and several billion pounds to do and is, in fact, the purpose of the international Human Genome Project. Instead, DNA from a few sites on the chromosomes is "chopped up" by certain enzymes which always go for particular features.
This gives rise to a set of fragments which would be the same sizes for DNA from the same person. These fragments are separated according to those sizes to give a pattern which should be unique for that person. But is it? This question has two aspects.
One relates to the reliability of the work carried out by DNA testing laboratories which, in some cases in the USA, was found to be rather sloppy. The other aspect relates to the assumptions behind DNA matching. ple, the chances of the DNA being identical at, say, four sites would be very small for close relatives and astronomically small for unrelated people.
But there is a very large assumption here. It is that the population of the world is thoroughly mixed. In practice, it is not and some communities are extremely close knit. The chances of two people having the same DNA at chosen sites start to increase.
The question is "Do the chances of a random match become high enough to cause serious doubt as to the identity of a criminal if there is no other evidence?" In some cases in the USA and Australia, the answer has been "Yes" and people have been acquitted in these cases. The defence argument was that DNA fingerprinting was based on assumptions genetic mixing found in general populations and that it had not been shown that such assumptions were valid for minority populations such as Hispanic or Vietnamese ones.
The controversy has erupted into the scientific press with the publication in the US journal Science of a paper by Richard Lewontin and Daniel Hartl highly critical of the claims of the DNA testers. Lewontin has previously been prominent in arguing against those who claim that I scores measure "intelligence" and show that intelligence is largely inherited.
The article argues that the DNA test methods are "liable to potentially serious errors". Contrary to the assumption of genetic mixing in ethnic groups, they say that these rouDs are made up of "multiple subpopulations that are genetically diverse". This leads to a far higher chance that two individuals from one of these subgroups would share the same pattern.
In addition, they say that the statistical methods currently used are inadequate. Unusually, the editor of Science invited defenders of DNA testing to reply in the same issue. In addition, it seems that the FBI tried to get the paper withdrawn before publication.
It would seem that much more research into the genetics of populations will be needed before DNA fingerprints can be accepted "South Africa at the crossroads"
SW London SO meeting, 8.00, Lambeth Town Hall, Brixton Saturday 25 January Bloody Sunday demonstration, E1.50 plus 38 pence påp From PO Box 823, London SE15 4NA.
----- pull-quotes -----
SCIENCE COILUIMN
without question in all cases. Wednesday 22 January
Socialists answer the New Right