The Genetic Barcode
In 1984, Alec Jeffreys at the University of Leicester discovered that certain regions of human DNA vary enormously between individuals. These hypervariable regions — now analyzed as Short Tandem Repeats (STRs) — act like a biological barcode. By measuring the number of repeats at multiple independent loci, forensic scientists create a profile so specific that the probability of two unrelated people sharing it is typically less than 1 in a trillion.
The Product Rule
The statistical power of DNA fingerprinting comes from independence. If each STR locus is inherited independently (which they are, being on different chromosomes), the probability of a full-profile match equals the product of individual locus match probabilities. Even if each locus has a modest 1-in-100 match probability, 13 independent loci yield 1 in 100¹³ — a number so vast it dwarfs the number of humans who have ever lived.
Population Genetics Matters
Match probability calculations depend on allele frequencies within the relevant population. Allele frequencies differ between ethnic groups due to population history. Forensic labs maintain frequency databases for major population groups and typically report the most conservative (highest) match probability. The NRC II report established guidelines for handling population substructure in court.
Limitations and Controversies
DNA statistics assume the evidence sample is clean, single-source, and properly handled. In practice, crime scene DNA is often degraded, mixed with multiple contributors, or present in trace quantities. Mixture interpretation — separating two or more contributors — remains an active area of research and courtroom debate, leading to the development of probabilistic genotyping software.