For this reason, random variations crop up in the non-coding (junk) DNA sequences as often as once in every 200 DNA letters. DNA fingerprinting takes advantage of these changes and creates a visible pattern of the differences to assess similarity.
Stretches of DNA can be separated from each other by cutting them up at these points of differences or by amplifying the highly variable pieces. ‘Bands’ of DNA are generated; the number of bands and their sizes give a unique profile of the DNA from whence it derived. The more genetic similarity between a person – or grape – the more similar the banding patterns will be, and the higher the probability that they are identical.