Sequoia

Sequoia analyzes Oxford Nanopore direct RNA long-read electric-current signals to extract features and identify RNA modifications such as m6A and m5C at single-molecule resolution.

Key Features:

• Multi-view visualization: Provides multi-view visual analytics of nanopore electric-current signals and sequence-level features.
• Python-based backend: Implements data processing and integration using a Python backend.
• Data import and processing: Imports raw Oxford Nanopore Fast5 files from direct RNA sequencing and processes signal-level data.
• Sequence clustering: Clusters reads based on electric-current signal similarities to group characteristic signal patterns.
• Signal exploration and feature identification: Enables iterative visual exploration and tuning of dimensionality reduction parameters to distinguish modified from unmodified RNA sequences.
• Discovery of signal signatures: Facilitates identification of qualitative signal signatures that characterize RNA modifications such as m6A and m5C.

Scientific Applications:

• RNA modification analysis: Supports detection and characterization of post-transcriptional modifications (m6A, m5C) from direct RNA nanopore data.
• Feature discovery and classifier development: Enables discovery of signal features and supports development of automated classifiers to separate modified and unmodified RNAs.
• Nanopore direct RNA workflows: Complements computational workflows for Oxford Nanopore long-read direct RNA sequencing datasets and aids hypothesis generation about RNA dynamics.

Methodology:

Imports raw Fast5 sequencing data, clusters reads based on electric-current signal similarities, applies dimensionality reduction with parameter tuning for visual exploration and feature identification, and was applied to approximately 500,000 direct RNA reads from human HeLa cells focusing on m6A and m5C.