Long-read sequencing transcriptome quantification with lr-kallisto.

Area of Science:

• Genomics and Bioinformatics
• Molecular Biology
• Transcriptomics

Background:

• High-throughput short-read sequencing enables routine gene-level RNA abundance quantification.
• Accurate and affordable quantification of full-length transcript isoforms remains a significant challenge.
• Transcript isoform complexity and genetic variation complicate quantification efforts.

Purpose of the Study:

• To address the challenge of definitive full-length transcript isoform quantification using long-read sequencing data.
• To develop and validate a computational tool for accurate long-read RNA quantification.
• To assess the impact of exome capture on the accuracy of long-read quantification.

Main Methods:

• Development of lr-kallisto, an adaptation of kallisto for long-read sequencing data.
• Utilized Oxford Nanopore (ONT) sequencing data for method development and testing.
• Incorporated exome capture techniques to evaluate improvements in quantification accuracy.

Main Results:

• Demonstrated that fast and accurate quantification of long-read RNA sequencing data is feasible.
• Showed that exome capture significantly improves the accuracy of long-read data quantification.
• lr-kallisto successfully adapts existing quantification methods for long-read technologies.

Conclusions:

• Long-read sequencing technologies, coupled with lr-kallisto, can enable routine definitive isoform quantification.
• Exome capture is a valuable strategy for enhancing the precision of long-read transcript quantification.
• This work overcomes key bioinformatic challenges in long-read data analysis for transcript isoform resolution.