Two-dimensional graphical representation of DNA sequences and intron-exon discrimination in intron-rich sequences.

We have shown earlier that analysis of DNA sequences using a two-dimensional graphical representation provides considerable information on new global sequence patterns and homologies, repeated structures, relative base abundances, probable evolutionary paths and evolutionary divergence. We have also reported that at a more micro level the graphical representation reveals distinct differences in the features of intron and exon segments of eukaryotic sequences. In this paper, the distinguishing features of the intron and exon segments are exploited to show, through several examples of different gene structures, that an averaging procedure over the slopes of the representative maps provides an easy technique to differentiate between probable intron and exon regions. We thus expect that this method will enable a rapid search and preliminary indication of possible locations of protein coding regions in long eukaryotic sequences.