Original URL: http://www.theregister.co.uk/2006/12/27/forensics_feature/

Forensics reaches into the future

How far should we let the DNA database go?

By Christopher Williams

Posted in Applications, 27th December 2006 07:02 GMT

We began researching this piece long before the recent murders in Suffolk. That horror has served to remind us that forensic technologies are now often the first thought in any criminal investigation. The shiny power of DNA technology is in no doubt, but are we in danger of being dazzled by it?

The state of the art

The National DNA Database (NDNAD) was set up in 1995 by the then Conservative government. Eleven years on it is the largest repository of human genetic information on the planet.

It is run under a Home Office contract by the Forensic Science Service and governed by a tri-partite board comprising representatives from the Home Office, Association of Chief Police Officers, and Association of Police Authorities. Representatives from the government's genetic ethics advisory body the Human Genetics Commission are invited to sit in.

At the moment, the NDNAD stores DNA profiles. This data is not the raw sequences of DNA, instead it is reliant on a special species of DNA sequence present in every chromosome: tandem repeats. These are short stretches of up to 100 DNA "letters" which are repeated as if the sequence is stammering.

Because of the imperfect way DNA is copied and mixed up when we are making sperm or eggs, the number of copies of each tandem repeat is variable. This evolutionary quirk provides a neat shorthand for forensic scientists. While each variation of a tandem repeat may appear in a large proportion of the population, the additive effect of comparing their length at several locations (the FSS uses 20) is to uniquely identify each individual, even within families.

The technique was developed by Alec Jeffreys at the University of Leicester in the late 1970s, who was knighted for it in 1994.

DNA fingerprinting briefly breaks down like this: enzymes cut out the repeated stretches to be examined from the DNA and an electric current is applied to the mixture. Because of the negative charge on the DNA molecule, the fragments move with the electricity, and the lower the number of copies of a repeat, the faster it will move. Dyes are used to produce the famous banding pattern of a DNA profile.

Police have powers to take a buccal swab – from the inside of the cheek – from everyone they arrest, including children. And since January 2006, a change in the law means that every offence is now arrestable; there was a time when litterbugs did not risk a night in the slammer.

The CSI effect

The success of forensics-based dramas like CSI has ensured there's no shortage of applicants to study forensics. Hollywood and its public are enamoured by the supernatural potency of the discipline.

The ludicrous feats of deduction often scored by the stars of these shows have convinced many that modern forensics is fast-paced, glamourous work. But not so. The science backing up the evidence is becoming increasingly involved, expanding to include bioinformatics for genetic sequence data and chemistry and physics, as the limits of what molecules can tell us are tested.

Earlier this year at an event jointly run by the Home Office and the Engineering and Physical Sciences Research Council (EPSRC) we saw a project led by Cranfield University to replace the time-honoured tradition of experts making highly educated time of death guesstimates. An arsenal of analytical techniques is being developed to turn the art into a science.

Project leader Dr Peter Zioupos said: "The future of age-at-death determination lies with analytical techniques rather than expert-based assessments. Our technique could help to identify victims of accidents, natural disasters and suicide, as well as murder. We aim to offer it as a forensic service within the next 12 months."

Perhaps we'll see CSI's impossibly well-groomed investigators replaced by banks of whirring mass spectrometers. Or not.

Frakenstein's monster?

While everyone else seems permanently impressed with DNA forensics, the man who pioneered DNA fingerprinting has repeatedly voiced fears over the mission creep that has characterised the NDNAD's first decade. A series of law changes has seen the Home Office's power to take and retain profiles have fewer and fewer limits.

In autumn, Alec Jeffreys told the BBC:" When the DNA database was initially established it was to database DNA from criminals such that if they reoffended they could be picked up. Hundreds of thousands of entirely innocent people now populate that database. My view is that that's discriminatory, that those people will be very skewed socioeconomically and ethnically so there's clear discrimination there."

Thirty-seven per cent of all black men in the UK are on the database, compared to 10 per cent on white men.

The government revealed the full extent of the NDNAD's innnocents roster only recently, bumping it up to 1,139,445 in response to a parliamentary question. The new figure is eight times a total given earlier in the year.

Such civil libertarian collywobbles do not wash with the Prime Minister, who on a visit to the FSS said there should be "no limits" to the size of the NDNAD, and has volutarily submitted his own genetic profile.

There are murmurs of government complacency on the issues from within the forensic community; it is left to NGOs like Genewatch UK (which has an essential backgrounder here) and the Nuffield Council on Bioethics, which is consulting on the NDNAD and expects to report next autumn, to examine the controversy.

In his comments, Jeffreys also singled out for criticism the lack of parliamentary debate over familial searches, which hugely extend the database's reach.

Familial searching may indeed be one technology which will help satisfy Blair's ambitions for an all-encompassing DNA repository. It means that if authorities turn up a DNA profile from a crime scene and run it on the database, it's not a wasted enterprise if they don't turn up an exact match. Though our DNA fingerprint is effectively unique, it shares similarities with blood relatives'.

The FSS added familial searches to its armoury this year, along with DNAboost, a less controversial procedure which gives forensic scientists stronger tools at the other end of the DNA profile: the crime scene. Rather than an insufficiently enormous database, difficulty getting good samples from a crime scene is often the reason forensics draw a blank. DNA may be mixed from several individuals or have degraded over time.

DNAboost is a computer program the FSS developed inhouse which turns the problem of poor sample on its head. FSS consultant forensic scientist Dr Tim Clayton, who works with DNAboost, told us the lateral thinking at its foundation is "beautifully simple, like all the best ideas".

It works by turning the database search into a process of elimination, so rather than looking for a match, it compares the sample's DNA fingerprint to every entry in the NDNAD and ranks them for similarity. A lot of the time it ends up finding the person they're looking for, and we learned that there are already several active prosecutions which used DNAboost as part of the investigation.

It's not an evidence-building tool then, but an investigatory one, and so part of the mission creep which Alec Jeffreys fears. At the EPSRC event in March it was clear that that is the future of forensics technologies, however.

Improvements in current techniques are perhaps the best hope civil libertarians have to stymie further expansion of the NDNAD. So-called "lab-on-a-chip" analysis looks set to revolutionise collection and testing of crime scene samples. The "chips" at the centre of the putative devices are not silicon microchips, rather pieces of plastic an inch or so across with drilled capillaries. The chemicals needed for DNA analysis can be controlled using the capillary design, putting the power of DNA analysis in the hands of scene of crime officers.

The field is moving rapidly, and the research councils are keen to fund projects as the designs migrate from academic research into working prototypes. The EPSRC gave £721,000 to lab-on-a-chip efforts at the University of Hull, which expects to begin field trials of a device within two years. Their machine will aim to detect and separate out DNA ready for fingerprinting in double quick time.

The final frontier

Without more oversight, few people we spoke to see an end to the NDNAD's mission creep. It is only a matter of time before DNA sequencing is both fast and cheap enough for entire genomes to be mapped to order.

TheX Prize Foundation, which drove the race for the first commercial space flight, has already laid down the gauntlet by offering $10m for the first team to sequence 100 genomes in 10 days. The billionaires reckon the achievement will herald a new era of personalised medicine and exotic targeted therapies.

The argument for whole genome forensics is sure to follow. Cast-iron evidence will be the call; cases have been lost because of the small chance of a false positive match with DNA fingerprinting. The Home Office hangs on to DNA samples once it has fingerprinted them already, so anyone who develops the technology to sequence them could have a fat government contract awaiting them too - 3 million-plus samples await. ®