Machine Learning-Assisted Direct RNA Sequencing with Epigenetic RNA Modification Detection via Quantum Tunneling

Area of Science:

• Molecular Biology
• Bioinformatics
• Genomics

Background:

• RNA sequence analysis is crucial for disease diagnosis and understanding infections.
• Current sequencing methods lose RNA modification data, vital for cancer research.
• RNA modifications play a role in cancer evolution and disease progression.

Purpose of the Study:

• To develop a machine learning-assisted direct RNA sequencing method.
• To enable simultaneous detection of RNA sequences and RNA modifications.
• To improve RNA analysis for disease diagnostics and research.

Main Methods:

• Employed a single-molecule, long-read, label-free quantum tunneling sequencing technique.
• Integrated machine learning (ML) algorithms for data analysis and classification.
• Utilized Shapley additive explanations to interpret ML model findings.

Main Results:

• Achieved 100% classification accuracy for RNA decoding.
• Reached 98% accuracy for simultaneous decoding of RNA and RNA modifications.
• Demonstrated high selectivity and sensitivity in recognizing RNA and its modifications.

Conclusions:

• The developed ML-assisted direct RNA sequencing method is effective for molecular-level RNA analysis.
• This approach can decode both RNA sequences and epigenetic modifications simultaneously.
• Represents a significant advancement for RNA sequencing and its applications in diagnostics and research.