Method of DNA Analysis Using the Estimation of the Algorithmic Complexity

The computational approach has a major impact in current biomolecular research. The computation of the algorithmic complexity (Kolmogorov) is a recently introduced method of global analysis for DNA sequences. The complexity is not directly computable, but one can estimate it using the compressibility of the given sequence. Intuitively, the more compressible a sequence is, the less complex it is; a random sequence is virtually incompressible. We have developed a set of programs which work together with GenCompress (Chen, Kwong, Li – 2001).We have used them for analyzing a variety of sequences: complete viral genome, chloroplastic genome, eukaryotic genes and fragments of genes, etc. We have found that exons are virtually incompressible, as opposed to introns, which posess various compressibility ratios. This confirms the existence of long-range correlations within introns, but not in exons. The method can also be used for detecting low complexity sequences, as well as homogeneous domain-like regions. This new procedure has also enabled the construction of phylogenetic trees. As a conclusion, the study of the complexity of DNA sequences is a powerful tool in genome analysis. 53 http://lejpt.academicdirect.ro Method of DNA Analysis Using the Estimation of the Algorithmic Complexity Ioan OPREA, Sergiu PAŞCA, and Vlad GAVRILĂ