Exploring functional conservation in silico: a new machine learning approach to RNA-editing.

Area of Science:

• Genomics and Molecular Biology
• Bioinformatics and Computational Biology

Background:

• Molecular biology has long studied gene function regulation through mechanisms like gene duplication and expression control.
• RNA editing, a process altering single nucleotides in RNA, significantly increases transcriptome and proteome complexity.
• Understanding RNA editing is crucial for deciphering gene function and its evolutionary adaptations.

Purpose of the Study:

• To develop and apply a novel, AI-powered approach for assessing the functional conservation of RNA editing targeting mechanisms.
• To investigate the evolutionary conservation and divergence of RNA editing processes using computational methods.
• To leverage big data and artificial intelligence for a deeper understanding of RNA editing's role in biological complexity.

Main Methods:

• Utilized two AI learning algorithms: random forest (RF) and bidirectional long short-term memory (biLSTM) neural networks with an attention layer.
• Integrated RNA-editing data from databases and variant calling from matched RNA-seq and DNA-seq experiments across species.
• Developed an in silico cross-testing analysis method to evaluate the conservation and divergence of RNA editing mechanisms.

Main Results:

• Successfully predicted RNA-editing events by analyzing both primary RNA sequences and secondary structures.
• The AI models demonstrated capability in identifying conserved and divergent patterns in RNA editing across different species.
• The cross-testing analysis provided a robust in silico framework for assessing the evolutionary dynamics of RNA editing.

Conclusions:

• The developed AI approach offers a powerful new tool for studying the functional conservation of RNA editing.
• This research enhances our understanding of how RNA editing contributes to transcriptome and proteome complexity across evolution.
• The findings lay groundwork for further investigation into specific RNA editing mechanisms, such as adenosine-targeting.