STR (Short Tandem Repeat) Analysis

Short tandem repeats (STR) are, as the name suggests, very short stretches of DNA that are repeated back to back in various locations throughout the human genome. Typically, the repeating sequence is just two, three, or four base pairs in length, and the number of copies found back to back is variable across a wide range. Unlike the DNA sequences in coding genes, there is no "correct" number of repeats for any specific STR in the genome; they are simply areas within the genome where variation is normal and healthy. For any specific STR, each person will have two copies, one that is inherited from their mother at conception, and the other that is inherited from their father.

STRs are helpful in forensic and paternity testing. Because there is a lot of natural variation for STRs, the chance of two people matching for the exact number of repeats on both inherited copies of the STR is fairly small. By combining analysis of many STRs across the genome, the probability of two people matching by random chance is extremely low.

For example, if the normal variation for certain STR (we will call it STR-A) is from seven to twenty copies, there will be fourteen different lengths that could be passed on from either parent. Since each person has two copies of STR-A, the total number of possible combinations would be 196 (14 x 14). If each of these different STR lengths had an equally probability (1/14), and every possible combination was equally likely, the chance of two people matching on the pattern of STR-A would be 1/196. Now, suppose that there are many other STRs available for study (eg., STR-B, STR-C, STR-D . . . STR-Z) each with the same probabilities we chose for STR-A. The likelihood of matching by random chance alone across these many markers would be found by multiplying the probabilities of each one individually. For two markers, the probability of a random chance match drops to 1/38,416. Adding a third marker drops the probability to 1/7,529,536. The addition of a fourth marker decreases the likelihood to less than one in 14 billion.

The discovery of STRs has greatly advanced forensic science by drastically reducing the chances that a suspect will be wrongly identified on the basis of forensic testing.

see also DNA fingerprint; DNA profiling; DNA sequences, unique.