VARUS: sampling complementary RNA reads from the sequence read archive.

Area of Science:

• Bioinformatics
• Genomics
• Computational Biology

Background:

• Large volumes of next-generation sequencing RNA data are publicly archived.
• This data is valuable for genome annotation and transcriptome assembly.
• Selecting appropriate data subsets from heterogeneous archives is challenging.

Purpose of the Study:

• To present VARUS, a software tool for automated selection, download, and alignment of RNA sequencing reads.
• To enable efficient genome annotation using archived RNA-Seq data.

Main Methods:

• VARUS utilizes the species' binomial name and genome to select runs and reads from NCBI's Sequence Read Archive.
• An online algorithm is employed to cover transcripts adequately under limited resources.
• Reads are randomly sampled from automatically chosen runs.

Main Results:

• VARUS achieved higher sensitivity and specificity with fewer downloaded reads than manual selection for most tested species.
• RNA-Seq data sampled by VARUS is suitable for fully-automatic genome annotation using BRAKER.
• Demonstrated effectiveness across twelve eukaryotic genomes.

Conclusions:

• VARUS automates RNA-Seq data selection and quality control for genome annotation.
• Enables fully automated genome annotation pipelines across multiple species without accuracy loss.
• Facilitates efficient and accurate large-scale genome annotation efforts.