In an article in this week's issue of the science journal Nature, the pair set out detailed criticisms of the way forensic experts present what appear to be huge odds against wrongful identification implied by the fingerprint technique.
David Balding and Peter Donnelly, of Queen Mary and Westfield colleges in London, say analysis of DNA - which carries the unique genetic material that makes an individual - is valuable in establishing innocence, but should not be used as the sole basis for conviction. 'The costs of previous misunderstandings in unsafe convictions are impossible to assess . . .' the pair conclude.
The two mathematicians say that DNA profiles can, in the absence of other evidence, leave a realistic chance that at least one other person than the suspect could be the criminal.
DNA fingerprinting relies on the fact that apart from identical twins, no two people have the same DNA genetic code in their cells. The technique uses 'probes' to pick out key features of this code in samples from the scene of a crime. If these features match those in the DNA fingerprint from a suspect, this is strong evidence that the samples came from that person.
Forensic scientists often describe to juries the extremely small probabilities of anyone but the suspect producing so strong a match with DNA taken from samples: one in several million is not an uncommon figure.
But instead of emphasising the slim chance of a DNA match being produced by an innocent person, courts should consider the probability of a person being innocent in spite of their having produced a match, they say.
To do this the court should weigh up all other evidence of innocence. 'I don't want to rule out the use of DNA fingerprinting, but would like to see the prosecution being more careful in explaining to juries how they should interpret DNA evidence,' Professor Balding said yesterday.
DNA analysis of blood, semen and other samples to produce a barcode-like readout was developed by Alec Jeffreys, Professor of Genetics at Leicester University, in 1985. It received widespread acclaim as the forensic breakthrough of the century after its first use in court in 1987 led to the conviction of Colin Pitchfork of murdering two children.
Throughout the late 1980s the technique gained popularity in courts, both in the UK and US, until uncertainties over laboratory procedures and questions over the assumptions over population statistics began to chip away at its credibility as evidence.
But Professor Donnelly's fresh approach to the value of DNA fingerprints helped convince the Court of Appeal late last year to quash the conviction of Andrew Deen for rape - the first UK challenge. The court ordered a retrial.
At Mr Deen's trial in 1990, it was claimed that the probability of a match was one in 700,000. At the recent appeal, a second opinion of this probability reduced it to one in 33.Reuse content