A high-resolution single-molecule sequencing-based Arabidopsis transcriptome using novel methods of Iso-seq analysis

Area of Science:

• Plant molecular biology
• Genomics
• Bioinformatics

Background:

• Accurate transcript annotation is crucial for gene expression analysis.
• Single-molecule long-read sequencing offers transcript integrity but faces accuracy challenges.
• Sequencing errors, RNA degradation, and incomplete cDNA synthesis impact accuracy.

Purpose of the Study:

• To present the Arabidopsis thaliana Reference Transcript Dataset 3 (AtRTD3).
• To introduce novel methods for accurate splice junction and transcription start/end site determination.
• To improve transcript quantification and differential expression analysis.

Main Methods:

• Development of novel computational methods for splice junction and transcription start/end site identification.
• Utilizing Iso-seq data for high-accuracy transcript structure determination.
• Stratified approaches to identify high-confidence sites and remove fragmentary transcripts.

Main Results:

• AtRTD3 contains over 169,000 transcripts, doubling existing datasets and including over 1500 novel genes.
• Seventy-eight percent of transcripts feature accurately defined splice junctions and transcription start/end sites.
• Analysis of an Arabidopsis cold response dataset demonstrated improved transcript expression profiling and identification of differential transcription start/polyadenylation site usage.

Conclusions:

• AtRTD3 is the most comprehensive Arabidopsis transcriptome available.
• It enhances precision in analyzing gene expression, alternative splicing, and transcription start/end site usage from RNA-seq data.
• The developed methods are broadly applicable to single-molecule sequencing analysis across species.