VarMeter: a prediction method for the impact of glycogene variants

Area of Science:

• Biochemistry and Molecular Biology
• Genetics and Genomics
• Structural Biology

Background:

• Glycans are vital for physiological processes, and defects in glycan-related genes (glycogenes) cause congenital disorders of glycosylation.
• Current understanding of glycogene function and disease links is limited by human-specific roles and the complexity of glycan modifications.
• Conventional methods struggle to fully assess glycan diversity and complexity, necessitating complementary approaches.

Purpose of the Study:

• To review current methods for predicting missense variant impact using structural information.
• To introduce VarMeter and VarMeter2, computational frameworks for predicting pathogenic variants.
• To highlight the utility of these tools in uncovering glycogene functions and disease associations.

Main Methods:

• Systematic assessment of human genetic variants, focusing on missense mutations.
• Utilizing protein structure prediction tools, including AlphaFold.
• Employing computational frameworks (VarMeter, VarMeter2) incorporating 3D structural parameters, pLDDT scores, and Mahalanobis distance analysis.

Main Results:

• VarMeter successfully predicted pathogenic variants in the ClinVar database.
• VarMeter2 demonstrated improved prediction accuracy by integrating AlphaFold pLDDT scores and Mahalanobis distance.
• The tools accurately predicted pathogenic variants for four key glycan-related proteins.

Conclusions:

• Structure-based variant interpretation tools offer a novel approach to understanding glycogene function and disease.
• VarMeter2 enhances the prediction of pathogenic variants, contributing to clinical genetic variation interpretation.
• These advancements facilitate the discovery of previously unrecognized glycogene functions and their disease relevance.