NucBreak

NucBreak detects structural errors in genome assemblies by analyzing alignments of correctly mapped reads to localize insertions, deletions, duplications, inversions, and inter- and intra-chromosomal rearrangements.

Key Features:

• Structural Error Detection: Locates structural errors in genome assemblies, including insertions, deletions, duplications, inversions, and inter- and intra-chromosomal rearrangements.
• Unique Methodology: Analyzes alignments of correctly mapped reads and leverages alternative read alignments rather than focusing on discordantly mapped or soft-clipped reads.
• Comparative Performance: In benchmarking with simulated and real datasets, demonstrated high sensitivity and a lower false discovery rate compared to Pilon, REAPR, FRCbam, BreakDancer, Lumpy, and Wham.
• No Error Annotation: Reports accurate localization of detected structural errors without providing annotations of the errors.
• Technical Implementation: Implemented in Python and uses Bowtie2 for read alignment.

Scientific Applications:

• Assembly accuracy assessment: Assessing and localizing structural errors to evaluate genome assembly quality.
• Structural variant detection: Detecting structural variants and rearrangements in genome assemblies.
• Downstream analysis reliability: Improving the reliability of downstream genomic analyses by identifying assembly errors.

Methodology:

Analyzes alignments of correctly mapped reads, leverages alternative read alignments instead of discordant or soft-clipped reads, performs read alignment with Bowtie2, and is implemented in Python.