iLncRNAdis-FB

iLncRNAdis-FB predicts associations between long non-coding RNAs (lncRNAs) and diseases by integrating multiple biological information sources into structured feature blocks for computational inference.

Key Features:

• Integration of Biological Information: Integrates multiple biological data sources into structured feature blocks that encapsulate relevant biological features.
• Convolutional Neural Network (CNN) Architecture: Constructs lncRNA and disease similarity matrices to generate three-dimensional feature blocks that serve as input to a convolutional neural network.
• Noise Reduction: Applies a supervised fusion approach to reduce noisy and irrelevant information, retaining pertinent features for prediction.
• Performance Superiority: Experimental results indicate higher predictive performance compared with other state-of-the-art lncRNA-disease association predictors.

Scientific Applications:

• Disease Mechanism Exploration: Identifying lncRNA-disease links to support investigation of molecular underpinnings of diseases.
• Drug Discovery: Prioritizing lncRNAs associated with diseases as candidates for molecular targeting in drug discovery.

Methodology:

Constructs lncRNA and disease similarity matrices, generates three-dimensional feature blocks from these matrices by fusing multiple biological information sources, and trains a supervised convolutional neural network on the feature blocks to predict lncRNA-disease associations.