Informational structure of genetic sequences

Department of Polymer Research, Weizmann Institute of Science, Rehovot 76100, Israel.

Abstract

Only about 1/20 of DNA of higher organisms codes for proteins, by means of classical triplet code. The rest is largely silent ("junk""? "selfish"?), with functions, if any, only speculated. The triplet code, however, is not the only code (message) carried by the DNA sequences. One could think of three levels of molecular communication, where the same sequence "talks’ to various biomolecules, while having, respectively, three different appearances: DNA, RNA and protein. Since the molecular structures and, hence, sequence-specific preferences of these are substantially different, the original DNA sequence has to carry simultaneously three types of sequence patterns (codes, messages), thus, being a composite structure in which one and the same letter (nucleotide) is frequently involved in several overlapping codes of different nature. This multiplicity and overlapping of the codes is a unique feature of the Gnomic language of the genetic sequences, not known for other languages. The overlapping coexisting codes have to be degenerate in various degrees to allow an optimal and concerted performance of all the encoded functions. There is an obvious conflict between the best possible performance of a given function and necessity to compromise the quality of the sequence pattern which encodes the function in favor of other patterns, overlapping with the given one. It appears that the major role of various changes in the sequences on their "ontogenetic" way from DNA to RNA and to protein, like RNA editing and splicing, or protein post-translational modifications is to resolve the conflits between overlapping messages. New data will be presented strongly indicating that the gene splicing is a device to resolve such conflict between the code of DNA folding in chromatin and the triplet code for protein synthesis.