Long read Structural Variation calling remains a challenging but highly accurate way to identify complex genomic alterations. To address this challenge, we developed Sniffles2, a successor to SV-detection method Sniffles. Sniffles2 increases both accuracy (e.g genotype, insertion sequence), speed (higher parallelism) and has more advanced functionality to further promote new insights into the organism or diseases. In the well-established GIAB benchmark, Sniffles2 outperforms its closest competitors (e.g. SVIM, CuteSV) both in terms of accuracy and efficiency.

Furthermore, Sniffles2 solves the problem of family to population level SV calling to produce fully genotyped VCF files by introducing a gVCF file concept. Sniffles2 population mode is designed to be more efficient and accurate at identifying SV across a collection of samples from familial trios and up to large-scale cohorts of thousands of individuals. Sniffles2 furthermore enables the detection of mosaic SV in bulk long read data. We used Sniffles2 somatic mode on the genomic data from the brain of a patient with multiple system atrophy, a rare sporadic neurodegenerative condition similar to Parkinson’s disease, with 55x WGS ONT coverage. We were able to identify multiple somatic SVs for this brain region. Thus overall demonstrating the utility and versatility of Sniffles2 to identify SV from mosaic to population levels.

Learning Objectives:

1. Discuss long read sequencing in complex genomic alterations.

2. Discuss structural variation in complex genomic alterations.

3. Discuss structural variation and its importance in human health/disease.