Splice_sim: a nucleotide conversion-enabled RNA-seq simulation and evaluation framework.

Area of Science:

• Bioinformatics
• Molecular Biology
• Genomics

Background:

• RNA sequencing (RNA-seq) techniques are used to study chemical RNA modifications.
• These modifications introduce nucleotide conversions, leading to mismatches in sequencing reads.
• Mapping biases of these modified reads to reference genomes are not well understood.

Purpose of the Study:

• To develop a computational tool for simulating RNA sequencing data with nucleotide conversions.
• To evaluate the mapping accuracy of different RNA-seq mappers under various modification scenarios.
• To identify and propose strategies for mitigating mapping biases in RNA modification studies.

Main Methods:

• Developed splice_sim, a splice-aware RNA-seq simulation pipeline.
• Introduced user-defined nucleotide conversions at specified frequencies.
• Created mixture models of converted and unconverted reads for realistic simulation.
• Calculated mapping accuracies per genomic annotation using simulated datasets.
• Evaluated state-of-the-art spliced-read mappers on mouse and human transcript data.

Main Results:

• Simulated RNA sequencing datasets under realistic conditions, including metabolic RNA labeling and RNA bisulfite sequencing.
• Quantified mapping accuracies of leading spliced-read mappers.
• Identified specific biases associated with nucleotide conversion reads.
• Demonstrated the impact of modifications on mapping performance.

Conclusions:

• splice_sim provides a robust platform for evaluating RNA sequencing mapping biases.
• Understanding and addressing these biases is crucial for accurate interpretation of RNA modification data.
• The study offers strategies to improve data analysis and prevent misinterpretation in RNA modification research.