Jumper enables discontinuous transcript assembly in coronaviruses.

Area of Science:

• Virology
• Computational Biology
• Genomics

Background:

• Viral gene expression in Nidovirales, including SARS-CoV-2, utilizes discontinuous transcription.
• This process, distinct from eukaryotic alternative splicing, is crucial for understanding viral replication and pathogenesis.
• Accurate transcript assembly is essential for analyzing viral gene expression patterns.

Purpose of the Study:

• To develop a novel computational method for assembling discontinuous viral transcripts from short-read sequencing data.
• To evaluate the performance of the new method against existing transcript assembly tools.
• To apply the method to analyze viral transcriptomes and identify drug responses.

Main Methods:

• Introduction of the DISCONTINUOUS TRANSCRIPT ASSEMBLY problem.
• Development of JUMPER, a maximum likelihood-based method accounting for varying transcript lengths.
• Application of JUMPER to short-read data from SARS-CoV-1, SARS-CoV-2, and MERS-CoV samples.

Main Results:

• JUMPER significantly outperforms existing methods in classical transcript assembly simulations.
• The method successfully identifies canonical transcripts present in reference transcriptomes.
• JUMPER predicts novel, non-canonical transcripts, validated by orthogonal analyses.
• Analysis of treated samples revealed viral drug responses at the transcript level.

Conclusions:

• JUMPER provides a robust approach for analyzing discontinuous transcription in Nidovirales.
• The tool enables the discovery of novel viral transcripts and insights into viral drug responses.
• JUMPER facilitates detailed transcriptomic analyses of viruses under diverse conditions.