Multifractal detrended cross-correlation analysis of coding and non-coding DNA sequences through chaos-game representation

• We propose a new approach combining CGR and 2D MF-X-DFA methods to analyze DNA sequences of unequal lengths.
• The multifractal characteristics of coding and non-coding sequence of eight prokaryotes are studied.
• The existence of strong multifractal behavior is observed in the nucleotide sequences.

We propose a new approach combining the chaos game representation and the two dimensional multifractal detrended cross correlation analysis methods to examine multifractal behavior in power law cross correlation between any pair of nucleotide sequences of unequal lengths. In this work, we analyzed the characteristic behavior of coding and non-coding DNA sequences of eight prokaryotes. The results show the presence of strong multifractal nature between coding and non-coding sequences of all data sets. We found that this integrative approach helps us to consider complete DNA sequences for characterization, and further it may be useful for classification, clustering, identification of class affiliation of nucleotide sequences etc. with high precision.