Identification of m⁶A-Associated RNA Binding Proteins Using an Integrative Computational Framework

Area of Science:

• Molecular Biology
• Bioinformatics
• Epigenetics

Background:

• N6-methyladenosine (m⁶A) is a prevalent mRNA modification crucial for regulating RNA functions.
• Previous research has identified several RNA binding proteins (RBPs) involved in m⁶A regulation through experimental methods.

Purpose of the Study:

• To develop and apply an integrative computational framework for identifying m⁶A-associated RBPs.
• To analyze reproducible m⁶A regions using RBPs' binding data and assess the framework's predictive power.

Main Methods:

• An integrative computational framework combining enrichment analysis and a Random Forest classification model was developed.
• The framework utilized RNA binding protein (RBP) binding data to analyze reproducible m⁶A regions from independent studies.
• Performance was compared against sequence-based prediction methods.

Main Results:

• Enrichment analysis identified known m⁶A-associated RBPs (e.g., YTH domain proteins) and proposed RBM3 as a potential m⁶A-associated RBP in mice.
• A significant correlation was observed between m⁶A-associated RBPs and protein expression levels, rather than gene expression.
• The Random Forest model demonstrated competitive performance and revealed m⁶A's role in repelling RBPs, suggesting inference of interactions beyond sequence level.

Conclusions:

• An effective integrative computational framework was designed for identifying known and potential m⁶A-associated RBPs.
• The study provides a computational approach to gain deeper insights into m⁶A modifications and their associated regulatory proteins.