A new method to cluster DNA sequences using Fourier power spectrum.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Accurate classification of genes and genomes is crucial for understanding evolutionary relationships.
• Existing methods for sequence comparison, such as multiple sequence alignment and alignment-free techniques, face challenges with sequence length variability and computational efficiency.

Purpose of the Study:

• To introduce a novel clustering method for classifying DNA sequences, genes, and genomes.
• To develop an efficient and accurate approach for determining evolutionary relationships between DNA sequences, regardless of length.

Main Methods:

• Constructing binary indicator sequences for each nucleotide in a DNA sequence.
• Applying Discrete Fourier Transform (DFT) to generate power spectra from these indicator sequences.
• Calculating mathematical moments from the power spectra to create multidimensional vectors for cluster analysis.

Main Results:

• The proposed method effectively classifies genes and genomes by analyzing evolutionary relationships.
• It demonstrates the ability to compare DNA sequences of varying lengths using power spectra and moments.
• Experimental results confirm the method's efficiency and comparable accuracy against established alignment-based and alignment-free techniques.

Conclusions:

• The novel clustering method offers an efficient and accurate tool for gene and genome classification.
• Its ability to handle sequences of different lengths and its speed make it a valuable advancement in bioinformatics.
• This approach provides a faster alternative to traditional methods for analyzing evolutionary relationships in DNA sequences.