Graphical Workflow System for Modification Calling by Machine Learning of Reverse Transcription Signatures

Area of Science:

• Epitranscriptomics
• Bioinformatics
• Molecular Biology

Background:

• RNA modifications are crucial for cellular functions and are studied using cDNA sequencing.
• Existing bioinformatics tools for analyzing modification data are limited.
• High-throughput sequencing data mining is essential for advancing epitranscriptomics.

Purpose of the Study:

• To develop a versatile, user-friendly graphical workflow system for RNA modification calling.
• To leverage machine learning for accurate and efficient modification detection.
• To provide quality assessment parameters for optimizing experimental and sequencing protocols.

Main Methods:

• A graphical workflow system integrating trimming, mapping, and postprocessing modules.
• Quantification of mismatch and arrest rates at single-nucleotide resolution.
• Machine learning algorithms for modification calling and quality assessment.

Main Results:

• The system enables precise quantification of modification signatures from cDNA data.
• Quality assessment parameters guide improvements in library preparation and sequencing.
• Automated bioinformatics workflows accelerate optimization cycles for modification calling.

Conclusions:

• The developed workflow system enhances the analysis of RNA modifications.
• It provides a robust platform for machine learning-based epitranscriptomic data mining.
• Automation facilitates faster experimental optimization and data interpretation.