kmerPyramid: an interactive visualization tool for nucleobase and k-mer frequencies

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Bioinformatics analyses frequently utilize nucleobase or k-mer frequency distributions in DNA/RNA sequences.
• Visualizing multi-dimensional frequency matrices is challenging.
• Existing methods lack effective visualization for sequence distribution patterns.

Purpose of the Study:

• Introduce the R-package 'kmerPyramid' for enhanced visualization of sequence data.
• Provide a novel method for displaying nucleobase or k-mer distributions.
• Facilitate the analysis of intra- and inter-species sequence variations.

Main Methods:

• Developed an R-package named 'kmerPyramid'.
• Implemented projection of sequence data onto principal component space.
• Created an interactive 3D pyramid visualization for sequence points.
• Pre-calculated k-mer frequency matrices for reference bacterial and viral genomes.

Main Results:

• 'kmerPyramid' displays sequences as interactive points in a 3D pyramid.
• Users can manipulate the pyramid (rotate, zoom) for detailed inspection.
• The package includes k-mer frequency data for approximately 2000 bacteria and 5000 viruses.
• Effective visualization of sequence differences at intra- and inter-species levels.

Conclusions:

• 'kmerPyramid' offers a powerful and intuitive visualization tool for sequence frequency data.
• The package simplifies the identification of patterns and differences in genomic sequences.
• Applicable to various bioinformatics tasks, including metagenomic and phylogenetic analyses.