Contrasting and Combining Transcriptome Complexity Captured by Short and Long RNA Sequencing Reads

Area of Science:

• Genomics
• Transcriptomics
• Bioinformatics

Background:

• RNA sequencing (RNA-Seq) using short or long reads is crucial for transcriptomic variation analysis.
• Long reads excel at isoform and repetitive region analysis, while short reads offer better coverage and accuracy.
• Optimal methods for quantitatively comparing these technologies and integrating their data remain underexplored.

Approach:

• Developed a pipeline to assess matched long and short reads data using diverse transcriptome statistics.
• Evaluated performance across datasets, algorithms, and sequencing technologies.
• Introduced MAJIQ-L, an extension of MAJIQ software, for unified transcriptome analysis from both data types.

Key Points:

• Short reads identified approximately 50% more splice junctions compared to long reads.
• 10-30% of splice junctions detected by short reads were missed by long reads.
• Long reads detected significantly more intron retention events, highlighting the value of data integration.

Conclusions:

• Combining long and short reads RNA sequencing provides a more complete transcriptome analysis.
• MAJIQ-L enables a unified view of transcriptomic variations from both technologies.
• The developed pipeline and software facilitate improved transcriptome analysis with future long-read technologies.