MapSplice

MapSplice detects splice junctions in RNA sequencing (RNA-seq) data to enable sensitive and specific identification of canonical and non-canonical splicing events.

Key Features:

• High sensitivity and specificity: MapSplice accurately identifies splice junctions and outperforms TopHat and SpliceMap on simulated RNA-seq data.
• Versatility with read lengths: Processes short reads (less than 75 base pairs) and long reads (greater than or equal to 75 base pairs).
• Independence from splice site features: Does not rely on predefined splice site features or intron length, enabling detection of novel canonical and non-canonical splice junctions.
• Utilization of read alignments: Leverages the quality and diversity of read alignments associated with a splice junction to improve mapping accuracy.
• Computational efficiency: Second-generation algorithm that maintains efficient CPU and memory usage while achieving high sensitivity and specificity.
• Benchmarking: Performance demonstrated and validated on simulated RNA-seq data.

Scientific Applications:

• Alternative splicing analysis: Detection of diverse splicing events for studies of gene expression regulation.
• Global splicing profiling: Recapitulates extensive alternative splicing on a global scale in experimental analyses.
• Cancer transcriptomics: Applied to RNA-seq from eight breast cancer datasets to highlight differences between molecular subtypes.

Methodology:

The algorithm maps reads spanning splice junctions, does not depend on traditional splice site features or intron length, and leverages the quality and diversity of read alignments associated with each junction.