ncRDeep

ncRDeep predicts non-coding RNA classes from sequence data using convolutional neural networks to support functional characterization and disease-related analyses.

Key Features:

• Convolutional Neural Network (CNN): Utilizes a CNN architecture to analyze RNA sequences and capture complex sequence patterns and dependencies.
• Sequence-based input: Relies solely on RNA sequence information and does not require RNA secondary structure predictions.
• Improved accuracy: Demonstrated an average accuracy improvement of 8.32% over state-of-the-art methods on benchmark datasets.

Scientific Applications:

• Functional Studies: Enables precise classification to support downstream functional analyses of non-coding RNAs.
• Disease Research: Aids identification of non-coding RNAs associated with diseases and cancers for biomarker or therapeutic target discovery.
• Biological Process Elucidation: Supports investigation of non-coding RNA roles in cellular mechanisms and biological processes.

Methodology:

ncRDeep integrates raw RNA sequence data into a convolutional neural network to extract predictive features without using RNA secondary structure predictions.