SlideSort-BPR

SlideSort-BPR detects chromosomal rearrangement breakpoints by performing reference-free comparisons of paired NGS short-read datasets to identify breakpoint reads and fusion transcripts.

Key Features:

• Reference-free comparison: Directly compares two sets of NGS short reads (normal versus potentially rearranged) without using a reference genome.
• All-against-all short-read algorithm: Implements a fast algorithm for all-against-all comparisons of short reads.
• Neighborhood-based theoretical analysis: Uses theoretical analyses of the number of neighboring reads to identify clusters of breakpoints.
• Breakpoint-read and fusion-transcript detection: Identifies breakpoint reads and predicts fusion transcripts from short-read data.
• Performance at 100× depth: Correctly identifies approximately 88% of breakpoints at 100× sequencing depth while reporting no false-positive reads in the reported evaluation.
• Benchmarking on cancer data: Evaluated on prostate cancer datasets and reported to predict more fusion transcripts accurately than previous methods while maintaining a low false-positive rate.

Scientific Applications:

• Cancer genomics: Detection of chromosomal rearrangements and fusion transcripts in cancer samples, including prostate cancer.
• Reference-free structural variant discovery: Comparative analysis of normal versus altered samples to discover breakpoint reads and structural rearrangements without a reference genome.
• Fusion transcript identification from NGS: Prediction of fusion transcripts directly from short-read NGS datasets.
• High-depth sequencing analysis: Breakpoint detection optimized for high sequencing depths such as 100×.

Methodology:

Performs direct reference-free comparisons of two NGS short-read datasets using a fast all-against-all short-read comparison algorithm combined with theoretical analysis of neighboring-read counts to identify clusters of breakpoints.