iFlnc

iFlnc identifies full-length long noncoding RNAs (lncRNAs) from human RNA-seq data to enable accurate annotation of novel and annotated full-length lncRNA transcripts.

Key Features:

• Comprehensive Identification: Identifies both novel and previously annotated full-length lncRNAs from RNA-seq data, including single-exon lncRNAs often missed by conventional methods.
• High Prediction Accuracy: Achieves prediction accuracy exceeding 85%, compared to under 50% reported for conventional approaches.
• Machine Learning Integration: Employs machine learning models that incorporate four feature types: transcript length, promoter signature, multiple exons, and genomic location.
• No Requirement for Additional Data: Operates solely on RNA-seq data and does not require transcriptional initiation profiling such as H3K4me3 ChIP-seq.
• State-of-the-Art Performance: Demonstrates an Area Under the Receiver Operating Characteristic curve (AUROC) greater than 0.92.

Scientific Applications:

• lncRNA annotation from RNA-seq: Provides reliable annotation of full-length lncRNAs directly from existing RNA-seq datasets.
• Regulatory role studies: Facilitates investigation of lncRNA regulatory functions in human development and disease.
• Single-exon lncRNA discovery: Enables detection and study of single-exon lncRNAs that are frequently overlooked by other methods.

Methodology:

Machine learning models integrate transcript length, promoter signature, exon structure, and genomic location features derived from RNA-seq data to classify full-length lncRNAs without relying on transcriptional initiation profiling such as H3K4me3 ChIP-seq.