Multifractal characterization of complete genomes

This paper develops a theory for characterization of DNA sequences based on their measure representation. The measures are shown to be random cascades generated by an infinitely divisible distribution. This probability distribution is uniquely determined by the exponent function in the multifractal theory of random cascades. Curve fitting to a large number of complete genomes of bacteria indicates that the gamma density function provides an excellent fit to the exponent function, and hence to the probability distribution of the complete genomes.