High-throughput annotation of full-length long noncoding RNAs with capture long-read sequencing

Area of Science:

• Genomics
• Transcriptomics
• Molecular Biology

Background:

• Accurate gene and transcript annotation is crucial for genomics.
• Current annotation techniques lack both high throughput and accuracy.
• Many gene models are incomplete, especially for long noncoding RNAs (lncRNAs).

Purpose of the Study:

• To accelerate lncRNA annotation using a novel technique.
• To reannotate lncRNA populations in human and mouse tissues.
• To characterize genomic features of lncRNAs, including protein-coding potential.

Main Methods:

• Development of RNA Capture Long Seq (CLS) by the GENCODE consortium.
• CLS combines targeted RNA capture with third-generation long-read sequencing.
• Experimental reannotation of GENCODE intergenic lncRNA populations in human and mouse tissues.

Main Results:

• CLS identified novel transcript models for 3,574 human and 561 mouse gene loci.
• CLS approximately doubled the annotated complexity of targeted loci.
• CLS outperformed existing short-read sequencing techniques in annotation.
• Full-length transcript models allowed definitive characterization of lncRNA genomic features.

Conclusions:

• CLS effectively addresses the bottleneck in transcriptome annotation.
• The method generates manual-quality, full-length transcript models at high-throughput scales.
• CLS significantly advances the accuracy and completeness of gene and lncRNA annotation.