Structural Variant Detection and Nanopore

“Comprehensive structural variant detection: from population to mosaic level” was the intriguing title for the session Fritz Sedlazeck from Baylor College of Medicine presented at London Calling 2022. Sedlazeck spoke about challenges of determining structure variants and their importance for learning about evolution, genomic disorders, their impact on regulation, and their impact on phenotypes. Sedlazeck emphasized that long reads provide more comprehensive information about structure variants because of access in repetitive regions and phasing information. They then explained that Sniffles2 has better performance in speed and sensitivity, including with lower coverage. The new version also has automatic parameter optimization. Sniffles2 was also used in the World Record of blood to report in seven hours! Sedlazeck also showed data of the increase in publications using long-read approaches and Sniffles2. The program has faster merging and more accurate genotyping. Sniffles2 was also used to resolve SVs in the MECP2 duplication syndrome. It was interesting to hear the limitations of Sniffles2 and vcf file format. Sedlazeck is part of two initiatives: Gregor is attempting to solve unsolved Mendelian diseases with Nanopore. The second one is All of Us, a large project sequencing one million people’s genomes and assessing two million with microarrays. Sedlazeck mentioned that Nanopore would be part of this program. New challenges and applications include using Sniffles2 to detect low abundance variations: Sniffles2 non-germline mode They are validating findings with Illumina sequencing and 600x coverage with Nanopore! They are adapting and expanding their approaches to do single-cell whole genome sequencing. They are optimizing the bioinformatics and by now likely have tested it in more samples. Structure variant detection in a single cell sounds unbelievable… and I imagine there could be many diagnostic and research applicaitons!

Structure with repetitive components.
How can Sniffles2 improve detection of structural variants? Photo by Pixabay on Pexels.com