Predicting Functional Effects of Synonymous Variants: A Systematic Review and Perspectives

Area of Science:

• Genomics and Bioinformatics
• Computational Biology
• Precision Medicine

Background:

• High-throughput sequencing advances precision medicine, necessitating accurate genome variant effect prediction.
• Synonymous single nucleotide variants (sSNVs) are as common as non-synonymous ones, yet their functional impact prediction remains challenging.
• A lack of experimentally validated sSNV data hinders the development of predictive computational methods.

Approach:

• Evaluated nine computational methods for predicting sSNV effects using observed and artificially generated variant sets.
• Analyzed variant distributions to infer evolutionary selection pressures, hypothesizing generated variants are enriched for deleterious effects.
• Investigated the relationship between observed variant frequencies and prediction scores to establish reliable effect prediction thresholds.

Key Points:

• Generated variants, assumed to be enriched for deleterious effects due to evolutionary selection, were significantly more often predicted as impactful than observed variants across all tested methods.
• Established predictor-specific thresholds for reliable sSNV effect predictions.
• Identified that variants scoring above these thresholds were disproportionately generated, supporting the hypothesis of deleterious enrichment.

Conclusions:

• Existing computational methods show varying capabilities in identifying severe sSNV effects.
• No current predictor can reliably identify subtle sSNV effects on a large scale.
• Further development is needed to improve the prediction of synonymous variant impacts in genomic data analysis.