A genomic distance for assembly comparison based on compressed maximal exact matches

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Genome assembly quality is traditionally assessed using metrics like contiguity, coverage, and accuracy.
• Existing methods often rely on sequence alignments, which can be computationally intensive and may miss certain types of similarity.

Purpose of the Study:

• To propose a novel, alignment-free genomic distance metric for evaluating genome assemblies.
• To introduce a method based on compressed maximal exact matches (MEMs) as a complementary quality assessment tool.

Main Methods:

• Utilizing maximal exact matches (MEMs), which are ungapped, perfect repeats.
• Employing the normalized compression distance, an information-theoretic measure based on finite-context models, to quantify sequence similarity.
• Applying the proposed metric to four human genome assemblies for validation.

Main Results:

• The proposed genomic distance effectively captures similarities and hierarchical structures within genome assemblies.
• Maximal exact matches provide insights into the assembly's composition and potential errors.
• The metric demonstrates sensitivity to variations introduced during genome sequencing and assembly processes.

Conclusions:

• The alignment-free genomic distance based on compressed MEMs offers a valuable addition to standard genome assembly quality metrics.
• This approach provides a new perspective on sequence similarity and assembly structure, complementing existing methods.
• The method has implications for understanding the impact of sequencing and assembly strategies on genomic data representation.