In the search for the low-complexity sequences in prokaryotic and eukaryotic genomes: how to derive a coherent picture from global and local entropy measures

• Published in: Chaos, solitons and fractals Vol. 20; no. 1; pp. 127 - 137
• Main Authors: Acquisti, Claudia, Allegrini, Paolo, Bogani, Patrizia, Buiatti, Marcello, Catanese, Elena, Fronzoni, Leone, Grigolini, Paolo, Mersi, Giuseppe, Palatella, Luigi
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-04-2004
• ISSN: 0960-0779; 1873-2887
• DOI: 10.1016/S0960-0779(03)00435-1

We investigate on a possible way to connect the presence of low-complexity sequences (LCS) in DNA genomes and the non-stationary properties of base correlations. Under the hypothesis that these variations signal a change in the DNA function, we use a new technique, called non-stationarity entropic index (NSEI) method, and we prove that this technique is an efficient way to detect functional changes with respect to a random baseline. The remarkable aspect is that NSEI does not imply any training data or fitting parameter, the only arbitrarity being the choice of a marker in the sequence. We make this choice on the basis of biological information about LCS distributions in genomes. We show that there exists a correlation between changing the amount in LCS and the ratio of long- to short-range correlation.