K-mer analysis of long-read alignment pileups for structural variant genotyping

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Accurate genotyping of structural variants (SVs) is essential for advancing genomics research and understanding genetic diversity.
• Existing SV genotyping tools often struggle with accuracy, especially in complex genomic regions.
• The need for rapid and precise SV genotyping methods is critical for large-scale genomic studies.

Purpose of the Study:

• To introduce kanpig, a novel computational method for accurate structural variant (SV) genotyping.
• To evaluate kanpig's performance against state-of-the-art SV genotyping tools using benchmark datasets.
• To assess kanpig's utility for multi-sample projects and its ability to genotype complex genomic loci.

Main Methods:

• Kanpig utilizes variant graphs and k-mer vectors to rapidly generate SV genotypes.
• The method was benchmarked on recent SV datasets, comparing its single-sample genotyping concordance with existing tools.
• Kanpig's performance was further evaluated on a cohort of 47 genetically diverse samples, focusing on complex loci.

Main Results:

• Kanpig achieved a single-sample genotyping concordance of 82.1%, substantially outperforming existing tools (average 66.3%).
• The method demonstrated high accuracy in genotyping complex loci, including neighboring SVs, in multi-sample analyses.
• Kanpig processed a single sample's 20x long-read data in just 43 seconds, compatible with PacBio and Oxford Nanopore reads.

Conclusions:

• Kanpig represents a significant advancement in SV genotyping, offering superior accuracy and speed.
• The method's effectiveness in complex genomic regions and multi-sample projects makes it valuable for diverse genomics applications.
• Kanpig's efficiency and compatibility with long-read sequencing technologies facilitate its application in large-scale genomic research.