NanoSplicer

NanoSplicer identifies splice junctions from raw Oxford Nanopore electrical signal ('squiggles') to improve accuracy of RNA splice junction calling.

Key Features:

• Direct Utilization of Raw Signal Data: Operates directly on raw Oxford Nanopore squiggle data without relying on matched short-read data or annotated splice junctions.
• Squiggle-Based Comparison and Probabilistic Scoring: Compares observed squiggles at potential splice junctions to candidate junction squiggles, measures squiggle similarity, and computes probabilities to select the most likely junction.
• Robustness to Basecalling Errors: Improves splice junction identification accuracy, particularly when basecalling errors are prevalent near junctions.
• Validation on Controlled and Biological Samples: Validated on synthetic mRNAs with known splice junctions and biological mRNAs from a lung-cancer cell line, demonstrating improved junction identification.

Scientific Applications:

• RNA Isoform Characterization: Enables accurate identification of splice junctions for resolving RNA isoforms from Oxford Nanopore data.
• Complex Transcriptome Analysis: Supports discovery of novel or rare splice variants in complex transcriptomes by leveraging raw signal information.
• Study of Disease-Associated Splicing: Facilitates investigation of alternative splicing in diseases such as cancer by improving detection of splice junctions from nanopore reads.

Methodology:

Analyzes raw nanopore squiggle data without external annotations or short-read alignments. Compares observed squiggles at potential splice junctions with candidate squiggles to evaluate similarity. Calculates the probability of each candidate junction based on squiggle similarity to identify the most probable splice junction.