Efficient Storage and Analysis of Genomic Data: A k-mer Frequency Mapping and Image Representation Method

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• k-mer frequency analysis is vital for DNA sequence studies but faces storage challenges due to high dimensionality.
• Existing methods for storing k-mer data are inefficient for large-scale genomic datasets.

Purpose of the Study:

• To develop a novel, lossless compression method for k-mer frequency data.
• To optimize storage and analysis of large biological sequence datasets.
• To enable efficient similarity analyses on genomic data.

Main Methods:

• Utilized Chaos Game Representation (CGR) to map k-mers to coordinates.
• Generated raster images from CGR components, representing k-mer frequencies.
• Partitioned and labeled CGR maps to create fractal-like image structures.
• Represented entire k-mer frequency sets as single images.

Main Results:

• Achieved file size reduction up to 16-fold compared to plain text and 3-fold compared to binary formats.
• Successfully performed alignment-free similarity analyses on whole genome sequences from 14 plant species using generated images.
• Demonstrated image reconstruction and k-mer frequency retrieval.

Conclusions:

• The CGR-based image compression method offers a fast and efficient solution for storing and analyzing k-mer frequencies.
• This approach facilitates efficient access and processing of large biological sequence datasets.
• The method shows significant potential for advancing genomic data analysis and comparative genomics.