m6Acomet

m6Acomet predicts functional annotations for individual N6-methyladenosine (m^6A) RNA methylation sites by analyzing RNA co-methylation networks derived from public human MeRIP-Seq datasets to elucidate epitranscriptome-level gene regulation.

Key Features:

• Functional annotation of m^6A sites: Systematically annotates human m^6A sites by examining network characteristics of RNA methylation profiles, associating 339,158 putative gene ontology functions with 1446 human m^6A sites.
• RNA co-methylation network construction: Builds an RNA co-methylation network from public human MeRIP-Seq datasets across 32 independent experimental conditions.
• Co-methylation pattern analysis: Identifies sites exhibiting co-methylated patterns at the epitranscriptome layer to reveal coordinated regulatory relationships.
• Guilt-by-association inference: Applies the guilt-by-association principle to propagate functional annotations through the co-methylation network.
• Comprehensive database of predicted functions: Provides a collection of predicted biological functions for individual m^6A sites based on network-derived associations.
• Validation and benchmarking: Validates predictions against a soft benchmark and demonstrates performance superior to random predictors.

Scientific Applications:

• Epitranscriptome functional annotation: Linking individual m^6A sites to gene ontology functions to support studies of the epitranscriptome layer of gene regulation.
• Mediator protein analysis: Facilitating investigations of mediator proteins (readers, writers, and erasers) and their roles in RNA methylation.
• Network-level regulatory studies: Enabling exploration of coordinated regulatory networks and co-methylation patterns involving reversible m^6A RNA methylation.
• Epigenetic regulation research: Supporting broader studies of epigenetic regulation mediated by m^6A modifications.

Methodology:

Constructs an RNA co-methylation network from public human MeRIP-Seq datasets across 32 independent experimental conditions and applies the guilt-by-association principle to analyze network characteristics and predict gene ontology functions for individual m^6A sites.