Comprehensive Benchmarking with Guidelines for Analyzing Transposable Element-derived RNA Expression.

Area of Science:

• Genomics and Bioinformatics
• RNA Biology
• Computational Biology

Background:

• Transposable element-derived RNAs (teRNAs) are increasingly recognized for fundamental and pathogenic roles in humans.
• Accurate identification and quantification of teRNAs are crucial but challenging due to a lack of standardized computational methods.
• Existing computational tools require rigorous evaluation to establish best practices for teRNA analysis.

Purpose of the Study:

• To benchmark the performance of 16 representative computational tools for teRNA identification and quantification.
• To evaluate tool performance across multiple levels: family, unit, exon, and transcript.
• To provide guidelines and an integrated pipeline for best practices in teRNA research.

Main Methods:

• Benchmarking involved 120 simulated and 60 real-world paired datasets (long- and short-read data).
• Performance evaluation focused on accuracy and resolution at family, unit, exon, and transcript levels.
• Development of decision-tree-style guidelines and an integrated best-practice pipeline.

Main Results:

• The exon level provides a trade-off between accuracy and resolution in teRNA analysis.
• Evaluated methods exhibit distinct strengths and weaknesses depending on the analysis level.
• The study establishes a framework for developing and benchmarking future computational tools.

Conclusions:

• This comprehensive benchmarking provides critical insights into the performance of teRNA identification tools.
• The developed guidelines and pipeline will aid researchers in selecting appropriate methods for functional studies.
• The evaluation framework serves as a gold standard for advancing computational teRNA analysis.