Bookend: precise transcript reconstruction with end-guided assembly.

Area of Science:

• Computational Biology
• Genomics
• Molecular Biology

Background:

• Accurate full-length transcript assembly is crucial for understanding gene expression and function.
• Existing RNA sequencing (RNA-seq) methods often struggle with precise identification of transcript start and end sites.
• Single-cell RNA-seq presents unique challenges for transcript assembly due to limited starting material and read distribution.

Purpose of the Study:

• To develop a computational package, Bookend, for improved transcript assembly.
• To leverage RNA 5' and 3' end information for enhanced transcript annotation.
• To demonstrate the utility of Bookend across various RNA-seq data types and biological systems.

Main Methods:

• Development of the Bookend software package for transcript assembly.
• Integration of data from diverse RNA-seq techniques, including short-read, long-read, and end-capture methods.
• Utilization of end-labeled reads from full-length single-cell RNA-seq datasets.

Main Results:

• Correct identification of transcript start and end sites is essential for precise full-length transcript assembly.
• Incorporating end-labeled reads significantly improves transcript assembly precision in single cells.
• Hybrid assembly of multiple RNA-seq datasets from Arabidopsis thaliana and meta-assembly from mouse embryonic stem cells yielded reference-quality annotations.

Conclusions:

• Bookend provides a robust framework for accurate full-length transcript assembly.
• The accurate identification and utilization of RNA 5' and 3' ends are critical for high-precision transcript annotation.
• Bookend's hybrid and meta-assembly capabilities enable the generation of comprehensive, reference-quality transcriptomes.