Characterizing exons and introns by regularity of nucleotide strings

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Numerical translation of DNA nucleotides aids in analyzing protein-coding and noncoding sequences.
• Previous methods revealed higher long-range correlations and regularity in intron-containing genes and regulatory DNA.
• Existing spectral analysis tools for DNA regularity are sensitive to nucleotide-to-number assignment rules.

Purpose of the Study:

• To propose and illustrate a novel spectral analysis method for DNA sequences.
• To develop a spectral tool invariant to nucleotide numerical translation rules.
• To analytically investigate DNA structural characteristics using numerical matrix representations.

Main Methods:

• Representing DNA nucleotide sequences as numeric matrices.
• Applying 2-dimensional wavelet transformation to assess sequence regularity via spectral slope.
• Calculating global and 'evolutionary' slopes for DNA sequences, exemplified on honeybee chromosome 1 gene sequences.

Main Results:

• Introns exhibit significantly higher regularity (more negative spectral slopes) than exons in analyzed honeybee genes.
• The proposed spectral tool's results are invariant to the specific numerical assignment of nucleotides.
• This invariance eliminates ambiguity present in traditional numerical translation methods for DNA analysis.

Conclusions:

• The novel spectral analysis method provides an objective measure of DNA sequence regularity.
• Intron regularity findings align with existing literature, validating the new approach.
• The invariance to translation rules represents a significant advancement for DNA sequence characterization.