374

28

Forensic Investigation

of genetic material. Volatile organic compounds are of course also fugacious but

unless they strongly adsorb somewhere they are permanently so, whereas any living

organism is constantly shedding fragments of skin (including microörganisms resi-

dent thereon), hair etc. apart from biofluids from lesions etc. such as blood, and the

DNA in these fragments or fluids is extremely involatile and fairly chemically stable

as well. The ability of the polymerase chain reaction (PCR) to repeatedly duplicate

DNA allows, in principle, even a single piece of DNA recovered (e.g., from a crime

scene) to yield sufficient material for sequencing analysis.

If two samples are being compared (e.g., one derived from the scene of the crime

and the other taken from a suspect) then if the distance between them (cf. Sects. 6.3.1

and 7.4.1) is zero, the samples can be said to match perfectly (with due regard for

noise in the system).

To avoid the laborious work of sequencing (and the inevitable uncertainties—to

some extent there is a trade-off between accuracy and rapidity), sequence informa-

tion can be used indirectly by making use of restriction endonucleases (REs), which

cut DNA at a place precisely defined by a short (4–8 bases) sequence. Thousands of

different REs are now known and hundreds are commercially available in purified

form. A sample exposed to a particular RE is fragmented in a characteristic and

completely reproducible manner. The distribution of fragment lengths, which can be

obtained by gel chromatography or capillary electrophoresis, is therefore a unique

(according to a specified precision—the abundance of a particular fragment is now

no longer a “digital” quantity because of the vagaries of chromatography for elec-

trophoresis) “fingerprint” of the sample. It is often much faster and cheaper to obtain

the fragment fingerprint than the actual sequence.

Problem. How many different DNA tetramers are there? How many octamers? What

are the abundances of the different tetramers in (a) a random DNA sequence of the

same length (as) and (b) the human genome; and what is the distribution of distances

between their occurrences?

Similar approaches are used to identify microörganisms used in biological ter-

rorism or warfare and their origin; and, very topically, the origin of microörganisms

causing epidemics and pandemics.

28.1

DNA Forensics in Criminal Investigations

It would be impracticable to sequence entire genomes; moreover the DNA recovered

from crime scenes is often damaged and the presence of entire intact genomes is

unlikely. Fortunately, the roughly 90% of the human genome that does not code for

proteins contains many repeated motifs, such as variable number of tandem repeats

(VNTRs) or short tandem repeats (STRs), which appear to be uniquely different

for each individual and, hence, can be used for identification purposes. Since these

noncoding repeats do not appear to be subjected to any selection pressure (at least,

not as far as is presently known—noncoding DNA may play a rôle in regulating gene