[A new approach to the study of statistical properties of genetic sequences]

An approach to the study of the properties of genetic texts is proposed. It is based on the investigation of the frequencies of all possible words (subsequences) in a text. The most important effect is that the original text could be reconstructed completely without deletions and/or mistakes using the set of words which are met in the text as a single copy. The length of words for which the effect occurs is a measure of the text redundancy. Some real genetic sequences were studied as well.