DNA sequence similarity analysis using image texture analysis based on first-order statistics.

Area of Science:

• Computational Biology and Bioinformatics
• Digital Image Processing
• Genomics

Background:

• DNA sequence similarity analysis is crucial for understanding evolutionary relationships, gene function, and protein structure.
• Traditional methods often rely on sequence alignment, prompting the need for novel alignment-free approaches.
• Developing new mathematical descriptors for DNA sequences is a key challenge in bioinformatics.

Purpose of the Study:

• To present a novel approach for DNA sequence similarity analysis using texture image processing.
• To explore the adaptation of texture analysis methods for alignment-free DNA sequence similarity calculations.
• To develop new mathematical descriptors for DNA sequences based on image texture.

Main Methods:

• DNA sequences were converted into gray-level textures by assigning values to nucleotides.
• Histogram-based texture analyses utilizing first-order statistics were employed for similarity calculations.
• Texture features were extracted from three different DNA datasets, and similarity matrices were computed.

Main Results:

• The phylogenetic trees generated by the texture analysis method showed similarity to those produced by MEGA software (sequence alignment-based).
• Texture analysis metrics effectively characterized DNA sequences.
• The approach demonstrated potential for alignment-free DNA sequence similarity analysis.

Conclusions:

• Texture analysis provides a viable method for characterizing DNA sequences and performing similarity calculations.
• This novel approach offers a promising alternative for alignment-free DNA sequence analysis in bioinformatics.
• The findings suggest that image processing techniques can be effectively integrated into computational biology studies.