A Novel Approach to Clustering Genome Sequences Using Inter-nucleotide Covariance

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Phylogenetic analysis of DNA sequences is crucial but often computationally intensive.
• Existing methods like Multiple Sequence Alignment (MSA) are time-consuming.
• Current alignment-free methods may lack accuracy due to manual intervention and neglect nucleotide interactions.

Purpose of the Study:

• To introduce a novel, efficient, and accurate method for DNA sequence classification and phylogenetic analysis.
• To address limitations of existing time-consuming and less accurate methods.
• To incorporate nucleotide interactions into sequence representation.

Main Methods:

• Development of the Accumulated Natural Vector (ANV) method.
• Representation of DNA sequences as points in a 18-dimensional space (ℝ¹⁸).
• Calculation of Accumulated Indicator Functions to derive ANVs, capturing nucleotide distribution and covariance.

Main Results:

• The ANV method accurately represents DNA sequences in ℝ¹⁸.
• It effectively captures both nucleotide distribution and inter-nucleotide covariance.
• Tests demonstrated the method's accuracy and time-efficiency across diverse datasets.

Conclusions:

• The Accumulated Natural Vector method provides an accurate and computationally efficient approach for DNA sequence classification.
• It enhances phylogenetic analysis by considering nucleotide interactions.
• ANV facilitates rapid global comparison of DNA sequences and genomes.