AlphaMissenseR: an integrated framework for investigating missense mutations in human protein-coding genes

Area of Science:

• Genomics
• Bioinformatics
• Artificial Intelligence

Background:

• Predicting missense mutation pathogenicity is crucial for understanding genetic diseases.
• Existing tools often lack comprehensive analysis and visualization capabilities.
• AlphaMissense offers AI-driven predictions for the entire human proteome.

Purpose of the Study:

• To introduce AlphaMissenseR, an R/Bioconductor package for accessing AlphaMissense predictions.
• To provide tools for analysis, visualization, validation, and benchmarking of variant pathogenicity.
• To facilitate integration with genomic analysis and clinical variant databases.

Main Methods:

• Development of an R/Bioconductor package (AlphaMissenseR).
• Integration with Bioconductor genomic region analysis tools.
• Implementation of multi-level visualization for genome browsers and 3D protein structures.
• Integration with clinical and experimental variant databases.

Main Results:

• AlphaMissenseR offers performant and reproducible access to AlphaMissense pathogenicity predictions.
• The package enables interactive exploration of variant effects across genomic regions and protein structures.
• Facilitates benchmarking of prediction methods against clinical and experimental data.

Conclusions:

• AlphaMissenseR enhances the utility of AlphaMissense predictions for researchers.
• The package supports comprehensive analysis, visualization, and validation of variant pathogenicity.
• It serves as a valuable tool for the bioinformatics and genomics communities.