Accurate identification and analysis of human mRNA isoforms using deep long read sequencing

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Accurate transcriptome analysis in eukaryotes is challenged by limitations of short-read sequencing.
• Current methods rely on aligning short complementary DNA reads to genomes, potentially missing novel structures.

Purpose of the Study:

• To evaluate the utility of long-read complementary DNA sequencing for comprehensive eukaryotic transcriptome analysis.
• To demonstrate an annotation-independent approach for identifying gene structures and splicing variations.

Main Methods:

• Generation of large-scale long-read complementary DNA datasets from human K562 and HeLa S3 cell lines.
• Annotation-independent alignment of long reads to infer gene structures.
• Assembly of long reads into full-length transcripts.

Main Results:

• Long-read alignments revealed partial gene structures consistent with annotations, identifying novel structures missed by short-read analysis.
• A higher fraction of novel gene structures was identified for long-noncoding RNA genes.
• Cell type-specific splicing patterns were accurately characterized without prior genome annotation.
• Successful assembly of full-length transcripts was achieved.

Conclusions:

• Long-read sequencing provides a robust, annotation-free method for eukaryotic transcriptome analysis, enhancing the discovery of novel gene structures.
• This approach is particularly valuable for analyzing transcriptomes of newly sequenced genomes and characterizing complex RNA species like lncRNAs.
• The method is broadly applicable across different long-read sequencing technologies.