DeepLncPro

DeepLncPro predicts long non-coding RNA (lncRNA) promoters from genomic sequences using an interpretable convolutional neural network (CNN) to enable identification of promoter-associated transcription factor binding motifs.

Key Features:

• Convolutional Neural Network Architecture: Employs a CNN to capture complex sequence patterns associated with lncRNA promoters.
• Interpretability: Extracts and analyzes transcription factor binding motifs from predicted lncRNA promoter regions.
• Cross-Species Validation: Validated on human and mouse genomes.
• Performance: Demonstrates higher accuracy than state-of-the-art machine learning methods and existing models for lncRNA promoter identification.

Scientific Applications:

• Regulatory mechanism elucidation: Identifies lncRNA promoters to support analysis of regulatory networks involving lncRNAs.
• Genome-wide promoter annotation: Enables computational genome-wide identification of lncRNA promoters.
• Transcription factor analysis: Provides motif-level signals to investigate transcription factor binding in lncRNA promoter regions.

Methodology:

Implements a convolutional neural network tailored for genomic sequence analysis and includes interpretable motif extraction to learn sequence patterns associated with lncRNA promoters and report transcription factor binding motif signals.