Ulysses

Ulysses detects low-frequency structural variations in large insert-size mate-pair sequencing libraries to improve identification and characterization of duplications, deletions, translocations, insertions, and inversions in repeat-rich genomic regions.

Key Features:

• High Detection Accuracy: Achieves higher detection accuracy than existing tools on simulated and real mate-pair sequencing datasets across duplications, deletions, translocations, insertions, and inversions.
• Statistical Significance Assessment: Assesses the statistical significance of candidate variants by comparing observed signals to an explicit model for experimental noise generation, enabling detection of low-frequency variants.
• Handling of Chimerical Sequences and Broad Insert Sizes: Accounts for broad insert size distributions and high rates of chimerical sequences common to mate-pair libraries to improve SV annotation.
• Mate-Pair Library Leveraging: Utilizes large insert-size mate-pair sequencing to provide higher physical coverage and access to repeat-containing regions of the genome.

Scientific Applications:

• Somatic Mosaicism Characterization: Detects low-frequency structural variants for characterization of somatic mosaicism in human tissues.
• Cancer Genomics: Identifies somatic structural variants in cancer genomes, including low-frequency somatic duplications as observed in tumor and blood samples from a breast cancer mate-pair library.

Methodology:

Compares observed variant signals to an explicit statistical model of experimental noise and processes large insert-size mate-pair sequencing data while accounting for broad insert-size distributions and chimerical sequences.