Between 2007 and
2009 police in Germany were frantically looking for a criminal. Although her
identity was not known, it was clear that she was dangerous: DNA evidence from
40 crime scenes, including several murder scenes, could be linked to her.
Despite immense
efforts lasting more than two years, they did not succeed. Eventually, they had to admit
that the reason why they had found DNA from the same woman in all those crime
scenes was that the cotton swabs used in the investigation were all
contaminated by the same cotton swab factory worker.
This episode of
forensic history illustrates how much law enforcement has come to rely on DNA
evidence. The centrepiece of this effort are national DNA databases, which
have acronyms like CODIS (USA), NDNAD (UK), or
FNAEG (France). Each of those databases contains millions of unique DNA
profiles.
Whilst DNA is central to modern forensics, DNA sequencing is not: The profiles in the databases are obtained using capillary electrophoresis of polymorphic loci called VNTRs. Is this likely to change with the arrival of more affordable sequencers?
Whilst DNA is central to modern forensics, DNA sequencing is not: The profiles in the databases are obtained using capillary electrophoresis of polymorphic loci called VNTRs. Is this likely to change with the arrival of more affordable sequencers?
Arthur Eisenberg,
who established one of the world's first forensics labs, thinks so. In a recent interview published on Genomeweb, he envisages a
future where all forensic DNA testing is done using sequencing.
More widespread use
of sequencing in forensics could have a number of benefits. Firstly, sequencing
would return information on DNA polymorphisms other than just VNTRs. This
includes SNPs,
which are much more common than VNTRs. SNPs are informative of traits like eye
colour, hair colour, and race, and could be used for facial composites
(although this would not be legal in some countries, including Germany).
Another advantage of
sequencing is that it can be applied to a wider range of forensic samples than
VNTR profiling. Shed hair is recovered from a lot of crime scenes, but cannot
be analysed using VNTR profiling because it usually does not contain nuclear
DNA. It does however contain mitochondrial DNA,
which is already used for forensic investigations.
A major challenge
for sequencing is backwards compatibility. Existing VNTR-based forensic
databases already have millions of entries, which are not easily comparable to
the output of most next generation sequencing (NGS) machines. The reason is
that NGS sequencers can typically only read 100-200 bases at a time, whilst for
gaining reliable information on VNTRs read lengths of 400 bases or more are
required. This is likely to become less of a problem in the next few years, as
read lengths increase.
Another remaining challenge is that NGS is not sufficiently reproducible to be admitted as evidence in a court of law. The same barrier exists for the widespread adoption of NGS in the clinic, where reproducibility is essential as well, but cannot yet be demonstrated for NGS-based tests either.
In summary, the hurdles for the widespread adoption of DNA sequencing in forensics seem to be technical, and solvable.
If you have a view
on whether sequencing has a future in forensics, please do not hesitate to post
your comment below.