Non-Synonymous and Synonymous Coding SNPs Show Similar Likelihood and Effect Size of Human Disease Association

Rong Chen, Eugene Davydov, Marina Sirota, Atul J. ButtePLoS 2010 · DOI

Abstract

Many DNA variants have been identified on more than 300 diseases and traits using Genome-Wide Association Studies (GWASs). Some have been validated using deep sequencing, but many fewer have been validated functionally, primarily focused on non-synonymous coding SNPs (nsSNPs). It is an open question whether synonymous coding SNPs (sSNPs) and other non-coding SNPs can lead to as high odds ratios as nsSNPs. We conducted a broad survey across 21,429 disease-SNP associations curated from 2,113 publications studying human genetic association, and found that nsSNPs and sSNPs shared similar likelihood and effect size for disease association. The enrichment of disease-associated SNPs around the 80th base in the first introns might provide an effective way to prioritize intronic SNPs for functional studies. We further found that the likelihood of disease association was positively associated with the effect size across different types of SNPs, and SNPs in the 3' untranslated regions, such as the microRNA binding sites, might be under-investigated. Our results suggest that sSNPs are just as likely to be involved in disease mechanisms, so we recommend that sSNPs discovered from GWAS should also be examined with functional studies.

Introduction

Genome-Wide Association Studies (GWAS) have identified thousands of genetic variants (SNPs) linked to hundreds of diseases. However, moving from a statistical association to a functional understanding of how a variant causes disease remains a significant challenge. Traditionally, researchers have prioritized functional validation efforts on non-synonymous SNPs (nsSNPs)—mutations that change the amino acid sequence of a protein—under the assumption that these are the primary drivers of disease. In contrast, synonymous SNPs (sSNPs), which alter the DNA sequence without changing the resulting protein, have often been dismissed as "silent."

This study challenges this bias, investigating whether nsSNPs actually possess a higher likelihood of disease association or a larger effect size than sSNPs and other non-coding variants. By analyzing over 21,000 associations, the authors demonstrate that synonymous mutations are just as likely to be associated with human disease and carry similar effect sizes as their protein-altering counterparts.

Methods and Data Curation

The researchers developed a comprehensive, manually curated database of human disease-SNP associations. The construction of this resource involved several key steps:

  1. Curation: Reviewing 2,113 publications from Medline/PubMed to extract 21,429 statistically significant associations.
  2. Feature Extraction: Recording over 100 features per association, including disease name, population, odds ratios (OR), p-values, and genetic models.
  3. Standardization: Mapping disease names to the Unified Medical Language System (UMLS) and annotating SNPs using the UCSC Genome Browser and NCBI Entrez to determine functional roles (e.g., intronic, intergenic, 3’-UTR).
  4. Statistical Analysis: Comparing the "likelihood" (percentage of a specific SNP type found to be disease-associated) and "effect size" (median odds ratio) across nine SNP categories.
Flowchart of the data curation and database construction process.

The study leveraged the HapMap 3 project as a reference set for baseline genomic distributions and data from six commercial genotyping platforms (Affymetrix and Illumina) to validate results across different technologies.

Key Results and Insights

The "Silent" Myth

The most significant finding is that sSNPs and nsSNPs share nearly identical likelihoods and effect sizes. In the HapMap 3 dataset, nsSNPs had a 1.46% likelihood of disease association, while sSNPs were nearly identical at 1.26%. Furthermore, the median odds ratio for nsSNPs was 1.72, compared to 1.70 for sSNPs (p = 0.24). This suggests that a synonymous mutation is generally as impactful as a non-synonymous one.

The 90% Gap

Focusing exclusively on amino-acid-changing variants (nonsense and nsSNPs) likely causes researchers to miss more than 90% of all disease-associated variants. While nonsense SNPs (creating premature stop codons) showed the highest likelihood (up to 16%) and median odds ratio (1.76), they represent only a small fraction of the total landscape.

Intron and UTR Findings

The study identified a specific enrichment of disease-associated SNPs around the 80th base of the first intron, providing a potential "rule of thumb" for prioritizing intronic variants. Additionally, SNPs in the 3’-untranslated regions (3’-UTR) showed high effect sizes but low discovery rates, suggesting these regions—often containing microRNA binding sites—are currently under-investigated.

Discussion and Limitations

While the study provides strong evidence for the importance of sSNPs, several factors must be considered. Linkage Disequilibrium (LD) remains a factor in GWAS interpretation; an sSNP might not be causal but could be "tagging" a nearby causal nsSNP. However, the authors cite examples of functional sSNPs affecting mRNA stability or protein folding rates to support their causal potential. Additionally, some associations might be "synthetic," where common variants appear associated due to LD with rare, high-impact mutations.

Future work is required to establish universal mechanisms for how sSNPs alter mRNA secondary structure, splicing, and translation kinetics. The paper concludes with a clear call to action: synonymous SNPs are not silent, and they are just as likely to hold the key to understanding human disease as mutations that change a protein's sequence.

Summary updated

Abstract

Many DNA variants have been identified on more than 300 diseases and traits using Genome-Wide Association Studies (GWASs). Some have been validated using deep sequencing, but many fewer have been validated functionally, primarily focused on non-synonymous coding SNPs (nsSNPs). It is an open question whether synonymous coding SNPs (sSNPs) and other non-coding SNPs can lead to as high odds ratios as nsSNPs. We conducted a broad survey across 21,429 disease-SNP associations curated from 2,113 publications studying human genetic association, and found that nsSNPs and sSNPs shared similar likelihood and effect size for disease association. The enrichment of disease-associated SNPs around the 80(th) base in the first introns might provide an effective way to prioritize intronic SNPs for functional studies. We further found that the likelihood of disease association was positively associated with the effect size across different types of SNPs, and SNPs in the 3' untranslated regions, such as the microRNA binding sites, might be under-investigated. Our results suggest that sSNPs are just as likely to be involved in disease mechanisms, so we recommend that sSNPs discovered from GWAS should also be examined with functional studies.
BibTeX
@inproceedings{chen2010non,
  author = {Rong Chen and Eugene Davydov and Marina Sirota and Atul J. Butte},
  title = {{Non-Synonymous and Synonymous Coding SNPs Show Similar Likelihood and Effect Size of Human Disease Association}},
  booktitle = {PLoS},
  year = {2010},
  doi = {10.1371/journal.pone.0013574},
}
Figure