DNA Fingerprinting

01 / UNDERSTAND & APPLY

1. Compare variable markers rather than reading an entire person

DNA fingerprinting, also called DNA profiling in modern usage, compares selected variable regions of DNA. A profile records features such as repeat-number alleles at several loci. It is not a photograph of chromosomes, a complete genome sequence or an image of the ridges on a finger.

Alec Jeffreys developed the original DNA-fingerprinting approach in 1984 through research on highly variable repeated DNA. Historical minisatellite band patterns and modern short tandem repeat profiles share the principle of comparing inherited variation, but their measurement methods and interpretation are not identical.

Most human DNA is shared among people. Profiling gains discriminating power by selecting loci with useful variation and combining information across them. One shared marker is unsurprising; a well-characterised multilocus profile can be much more informative. The strength of evidence must nevertheless be calculated within an appropriate model.

This lesson uses invented fragments, alleles and anonymous sample labels. Its purpose is to teach inheritance and evidence reasoning, not to identify an actual person, judge a legal case or draw conclusions about a real family. A matching profile does not by itself establish what someone did or when DNA reached a location.

02 / UNDERSTAND & APPLY

2. Repeat number creates measurable allele differences

A tandem repeat consists of a short sequence unit repeated beside itself. If one chromosome carries six copies and another ten copies at the same locus, the repeat region differs in length even if the repeat motif is the same. Variable number tandem repeat, or VNTR, is commonly used in school treatments for highly variable minisatellite regions. STRs have shorter repeat units and are widely used in modern profiling.

For a fictional repeat motif four base pairs long, an allele with eight copies contributes 32 bp of repeat DNA. Twelve copies contribute 48 bp. If the same measured flanking sequence is included in each fragment, the total fragment-length difference remains 16 bp. The flanks must not be counted as extra repeat units.

Polymorphism describes variation present in a population. Older teaching definitions often use a frequency threshold of about 1% to distinguish common polymorphisms from rarer variants. That convention is not a law stating that a DNA difference below the threshold cannot be inherited, detected or informative.

Repetitive DNA may occur in coding or noncoding contexts, but many identification markers are chosen outside protein-coding regions. Noncoding does not mean incapable of inheritance or automatically devoid of function. The word satellite historically relates to repetitive DNA fractions distinguished by physical properties; it does not mean a separate chromosome orbiting the nucleus.

03 / UNDERSTAND & APPLY

3. A diploid marker usually supplies two allele copies

At an ordinary autosomal STR locus, an individual normally carries one allele inherited through each parent. A genotype written 8,12 indicates repeat-number alleles eight and twelve; it does not mean twenty chromosomes or a single allele with twenty repeats. If both copies have eight repeats, the genotype is 8,8.

A heterozygote can yield two allele positions in a clean locus-specific length profile. A homozygote can yield one position because the same-length products coincide. One peak therefore does not imply that the person has only one chromosome or only one DNA molecule in the sample.

Across many autosomal loci, the pattern generally reflects Mendelian segregation. With a parent 8,12 and another 10,14, the possible child genotypes at that locus are 8,10; 8,14; 12,10; and 12,14 under the simple model. Each receives one allele from each parent.

These models assume correctly typed samples, ordinary diploid inheritance, no new mutation and no mixture or other complication. Real interpretation considers departures from those assumptions. A classroom mismatch under the model is an inference about that model, not an accusation or a definitive statement about a real relationship.

04 / UNDERSTAND & APPLY

4. Historical band patterns combined size separation with sequence recognition

In the classical VNTR approach, DNA fragments containing repeat regions were distinguished by length and detected with complementary probes. Conceptually, restriction cleavage produces fragments, electrophoresis separates them, denaturation makes strands available for hybridisation, and transfer to a membrane preserves their relative positions for probing.

A labelled probe binds a sufficiently complementary sequence. Detection then reveals selected fragments rather than every DNA fragment in the sample. Southern blotting concerns DNA; Northern blotting concerns RNA, and Western blotting detects proteins. Their names should not be exchanged merely because all can produce visible bands.

A multilocus probe can reveal many bands from different genomic locations, whereas a locus-specific interpretation concerns a defined marker. Counting every band in a historical multilocus picture as a separate allele at one diploid locus is therefore incorrect.

These are conceptual stages, not operating instructions. Probe choice, fragment integrity and specificity affect what is visible. A missing signal can reflect failure to detect material, so a diagram must state when it assumes complete, reliable detection.

05 / UNDERSTAND & APPLY

5. Fragment position is a measurement, not an identity verdict

DNA’s negatively charged backbone allows movement toward the positive electrode in an electric field. For comparable linear fragments in a suitable gel, smaller fragments usually travel farther than larger ones. A size standard is needed to relate positions to fragment lengths; the distance is not generally a simple linear ruler of base-pair number.

Modern STR profiling commonly targets defined regions with amplification and records the resulting allele positions instrumentally. A profile may be shown as peaks rather than dark bands. The position indicates an allele’s measured size or assignment; peak height is not a repeat count.

A fragment length includes the repeat region and any measured flanking sequence. If two loci have different flanks, identical total lengths do not necessarily mean identical repeat numbers. Locus identity and calibration are therefore essential parts of the result.

Sequence-based profiling can sometimes distinguish alleles that have the same overall length but different internal sequence. More detailed information also requires appropriate reference data and validation. There is no universal fixed number of loci that describes every historical method, jurisdiction and current platform.

06 / UNDERSTAND & APPLY

6. Compare corresponding loci and preserve the distinction between exclusion and support

In an idealised complete single-source profile, sample A has locus L1 genotype 8,12 and locus L2 genotype 10,11. Sample B has the same pair at each locus. They are concordant for the markers examined. Sample C has 8,13 at L1, so it differs at that marker even if L2 matches.

Alleles must be compared within the same locus. An allele labelled 12 at L1 does not compensate for a missing 12 at L2. Likewise, genotype 8,12 equals 12,8 because the order of the two unphased allele labels is not meaningful.

Under the stated clean single-source model, a reliable incompatible allele can exclude a proposed common source. A concordant profile supports compatibility but does not logically prove that no other individual could share it. Identical twins usually share the standard inherited STR profile, and relatives are more likely to share alleles than unrelated people.

An incomplete profile contains less information than a complete one. One should report the markers actually observed rather than silently filling gaps from a comparison sample. Additional data can strengthen or weaken an interpretation; uncertainty is part of the evidence, not a reason to replace it with certainty.

07 / UNDERSTAND & APPLY

7. Rarity depends on genotype frequencies and model assumptions

For an elementary autosomal locus in Hardy–Weinberg equilibrium, an allele with frequency p has homozygote frequency p². Two different alleles with frequencies p and q have heterozygote frequency 2pq, because the two parental transmission orders lead to the same unordered genotype.

If allele 8 has frequency 0.20 and allele 12 frequency 0.10, the idealised frequency of genotype 8,12 is 2 × 0.20 × 0.10 = 0.04. The frequency of 8,8 is 0.20² = 0.04 in this particular example. Equal numerical results do not mean the formulas are interchangeable.

For suitable independent loci, genotype frequencies can be multiplied to estimate a multilocus profile frequency. If three fictional locus-genotype frequencies are 0.04, 0.05 and 0.02, their product is 0.00004, or one in 25,000. The product rule requires the stated independence assumptions and appropriate population information.

Real forensic calculations can account for population structure, relatedness, sample uncertainty and other factors beyond this elementary model. A school exercise should supply its assumptions. It is invalid to invent allele frequencies, multiply linked markers as independent without justification, or apply an unrelated-person model to a specified close relative.

08 / UNDERSTAND & APPLY

8. A random-match probability is not the probability of guilt

A random-match probability concerns how often an unrelated randomly selected individual from the stated population model would have the specified profile. It does not directly equal the probability that a particular person is the source, and it certainly does not equal the probability of innocence or guilt.

Likelihood ratios compare how probable the observed evidence is under two explicitly stated propositions. For a fictional evidence event E, if P(E | H1) = 0.80 and P(E | H2) = 0.02, the likelihood ratio is 40. This says the evidence is forty times as probable under H1 as under H2 in that model.

It does not say that H1 has probability 40 or 97.5%. A posterior probability would additionally require prior odds and a defensible model connecting all the evidence. Swapping P(E | H) with P(H | E) reverses the conditional question.

Even a strong source-level result does not determine the activity that deposited DNA. The same biological material can be consistent with different timing or transfer circumstances. A responsible answer separates the measured marker pattern, the strength of source evidence and any broader conclusion about events.

09 / UNDERSTAND & APPLY

9. Mixed and low-level samples require additional uncertainty models

A sample can contain DNA from more than one contributor. Several individuals can share alleles, and their combined signals overlap. Therefore, finding all of one candidate profile’s alleles somewhere in a mixture does not establish that the candidate contributed: other contributors can jointly supply the same set.

Low-level or degraded material may yield incomplete detection. Allele dropout means a real allele is not observed; additional signals may arise from contamination or analytical artefacts. Stutter products associated with repeat amplification can complicate peak interpretation. A small peak is not automatically a true additional contributor allele.

For a fictional mixture with observed alleles 8,10,12,14 at one locus, possible pairings include contributors 8,10 plus 12,14, or 8,12 plus 10,14. The allele list alone does not distinguish these decompositions. More loci and an appropriate model are needed.

Mixture interpretation uses validated biological and statistical assumptions rather than merely counting coloured peaks. This lesson’s clean-profile puzzles must not be transferred unchanged to complex mixtures. If the conditions required for a simple inference are absent, the correct response is to state that limitation.

10 / UNDERSTAND & APPLY

10. A result is only as useful as its sample context and interpretation

A profile must remain linked to the correct sample identity and handling record. Contamination, sample switching or poor-quality material can change the evidential meaning even when the underlying genetic principles are sound. Controls and traceable documentation address different failure modes from mathematical match calculations.

The usual nuclear DNA profile is broadly shared across a person’s nucleated tissues, but sample biology still matters. Mature human red blood cells lack nuclei, so nuclear DNA recovered from blood commonly comes from other cellular components such as white blood cells. Hair material with and without root-associated tissue can differ in the nuclear DNA available.

DNA profiling has uses in identification, biological relationship studies, conservation and other research. Each application requires consent or other appropriate authority, suitable reference material and careful interpretation. Genetic data can reveal information about relatives as well as the sampled individual.

Privacy and fairness require more than removing names from a data table. Access, retention, secondary uses and the possibility of re-identification all matter. This is a scientific discussion of evidence and responsibility, not legal advice or instructions for covert sample collection.

11 / UNDERSTAND & APPLY

11. Eight original worked reasoning cases

All samples, loci and frequencies are invented educational examples, not casework.

12 / UNDERSTAND & APPLY

12. Original practice with explained answers

Keep locus identity, detection assumptions and conditional probabilities explicit.