The transition of new scientific concepts into forensic evidence acceptable to the courts has rarely been a smooth one. The legal profession is by nature conservative, distrustful of anything new, and scientists have on occasion been over-confident in the interpretation of their data. When ABO blood grouping became available as a means of excluding paternity some courts at first refused to accept it and some men who could not have been the putative fathers were served with paternity orders.
So it is somewhat surprising that the courts, particularly in the United States, accepted DNA typing, also known as DNA fingerprints, almost at once. Perhaps they did so because fingerprints have long been accepted as being unique to an individual. In Britain DNA fingerprints have been used quite successfully in a number of cases but in the US they have not been an unqualified success. Part at least of this different experience is that in Britain multi-locus probes (MLPs) have been used whereas in the US single-locus probes (SLPs) have been favoured.
MLPs give a more unique DNA fingerprint than SLPs but their complexity can make them more difficult to interpret and they require larger amounts of DNA than SLPs. Since it is also believed that SLPs are technically more robust for use with dirty forensic samples it is hardly surprising that many laboratories have preferred to use them especially when the amount of DNA recovered at the scene of a crime has been small.
However, the drawbacks of SLPs together with poor quality control in some of the forensic laboratories has led to a number of recent court decisions in the US where DNA fingerprint evidence has been ruled out. Exaggerated claims of astronomically high chances against a match occurring by chance have also led to a lack of confidence in the forensic value of DNA fingerprinting. There is a distinct possibility that this potentially valuable means of proving the guilt or innocence of a suspect may be treated as ABO blood group evidence was in its early days.
What is clearly needed is a more substantial body of background data from which accurately calculated estimates of the chances of a match occurring between two individuals can be obtained and a critical examination of the laboratory techniques used so that DNA fingerprint data can be confidently presented to a court. Fortunately, and perhaps just in time, studies of this kind are now starting to be published.
Alec Jeffreys of Leicester University who was one of the originators of DNA fingerprinting has, with two colleagues, carried out a monumental study of 1702 paternity cases using MLPs (Jeffreys et al, 1991). They show well separated thresholds for declaring paternity (over 40 per cent band sharing) and non-paternity (less than 20 per cent band sharing) with no cases falling between. They also show that the probability of a match occurring at random is extremely low and that the identification of an individual using MLPs is very accurate.
The problem of the accuracy of SLPs used in forensics has been examined in a collaborative study between the forensic laboratories of the United States Federal Bureau of Investigation and the Royal Canadian Mounted Police (Budowle et al, 1991). Starting with the premise that DNA fingerprints are the best available technique for excluding a crime suspect they have developed a conservative statistical method designed to ensure that undue weight is not placed on a sample attributed to an accused individual who cannot be excluded. This is only a start, although an encouraging one, and much more work on these lines remains to be done.
Unlike the ABO story where the courts took their time in catching up with the scientific data it seems that research on DNA fingerprints is now catching up with its courtroom applications. Since the potential value of DNA fingerprints in showing the guilt or innocence of an accused person is considerable, the sooner this happens the better.
References:
Jeffreys et al, Amer. J. Hum. Genet. v 48, p824, 1991.
Budowle et al, Amer. J Hum. Genet. v 48, p841, 1991.
John Timson