BoostNano

BoostNano preprocesses raw signal data from Oxford Nanopore Technologies (ONT) direct RNA sequencing to improve read quality before basecalling for downstream transcriptomic and epitranscriptomic analyses.

Key Features:

• Adapter Segmentation and Trimming: Identifies and removes adapter sequences from raw ONT signal-level data to provide cleaner input for basecalling.
• Poly(A) Stall Detection and Correction: Detects polyadenylation-associated sequencing stalls (poly(A) regions) and addresses these events to improve read continuity.
• Transcription Trimming: Trims transcriptional artifacts from raw signal data that can interfere with accurate basecalling.
• Deep Learning Algorithms: Applies deep learning models to segment and trim adapter sequences, poly(A) stalls, and transcription artifacts in raw ONT signal data.
• Pre-basecalling Preprocessing: Performs preprocessing on raw signal-level ONT direct RNA reads prior to basecalling to enhance downstream analyses.
• Benchmarking and Performance: In a comparative study using synthetic RNA standards (Sequins), BoostNano produced mean poly(A) tail-length estimates within 12% of correct values.
• Comparison with Dorado: The same study reported Dorado had faster run times and a lower coefficient of variation and was recommended for integration with basecalling.

Scientific Applications:

• Transcriptomics: Preprocesses full-length ONT direct RNA reads to improve transcript structure and isoform analyses.
• Epitranscriptomics: Enhances signal quality for detection and analysis of RNA modifications from ONT direct RNA sequencing.
• Polyadenylation Dynamics: Facilitates analysis of poly(A) tail lengths and polyadenylation dynamics in RNA molecules.
• Poly(A) Tail-Length Estimation: Supports benchmarking and tail-length estimation workflows using synthetic standards such as Sequins.

Methodology:

BoostNano employs deep learning algorithms to segment and trim adapter sequences, poly(A) stalls, and transcription artifacts from raw ONT signal data prior to basecalling.