Functional and Structural Consequence of Rare Exonic Single Nucleotide Polymorphisms: One Story, Two Tales

Abstract

Genetic variation arising from single nucleotide polymorphisms (SNPs) is ubiquitously found among human populations. While disease-causing variants are known in some cases, identifying functional or causative variants for most human diseases remains a challenging task. Rare SNPs, rather than common ones, are thought to be more important in the pathology of most human diseases. We propose that rare SNPs should be divided into two categories dependent on whether the minor alleles are derived or ancestral. Derived alleles are less likely to have been purified by evolutionary processes and may be more likely to induce deleterious effects. We therefore hypothesized that the rare SNPs with derived minor alleles would be more important for human diseases and predicted that these variants would have larger functional or structural consequences relative to the rare variants for which the minor alleles are ancestral. We systematically investigated the consequences of the exonic SNPs on protein function, mRNA structure, and translation. We found that the functional and structural consequences are more significant for the rare exonic variants for which the minor alleles are derived. However, this pattern is reversed when the minor alleles are ancestral. Thus, the rare exonic SNPs with derived minor alleles are more likely to be deleterious. Age estimation of rare SNPs confirms that these potentially deleterious SNPs are recently evolved in the human population. These results have important implications for understanding the function of genetic variations in human exonic regions and for prioritizing functional SNPs in genome-wide association studies of human diseases.

Keywords: RNA structure; ancestral allele; positive selection; purifying selection; single nucleotide polymorphisms; translational selection.

Fig. 1.—

Correlation between MAF of exonic SNPs and corresponding conservation/pathogenicity score. (A) Relationship between MAF and PhyloP score; (B) Relationship between MAF and GERP score; (C) Relationship between MAF and CADD score. Each point represents the mean of the corresponding category. Error bars indicate the standard error of the mean.

Fig. 2.—

Functional effects caused by nonsynonymous SNPs as a function of DAF. (A) Amino acid chemical distance caused by nonsynonymous SNPs as a function of DAF; (B) Change in amino acid hydrophobicity index (|ΔH|) as a function of DAF; (C) BLOSUM62 scores as a function of DAF; (D) SIFT scores as a function of DAF; (E) Comparison of the functional effect caused by nonsynonymous SNPs between the genes undergoing stronger and weaker purifying selection. The median of ω (calculated using mouse orthologs) is used as the cutoff. Each point represents the mean of the corresponding category. Error bars indicate the standard error of the mean.

Fig. 3.—

The mRNA structural effect caused by exonic SNPs as a function of DAF. (A) Structural entropy caused by exonic SNPs as a function of DAF; (B) MFE gap between two alleles (|ΔΔG_MFE|) as a function of DAF; (C) Structural distance between two alleles as a function of DAF; (D) Comparison of the structural entropy between SNPs located in 5′- and 3′-UTR; (E) Comparison of the structural entropy between nonsynonymous and synonymous SNPs. Each point represents the mean of the corresponding category. Error bars indicate the standard error of the mean.

Fig. 4.—

Change in codon optimality (|ΔO_codon|) as a function of DAF. Each point represents the mean of the corresponding category. Error bars indicate the standard error of the mean.

Fig. 5.—

Cumulative distribution of DAF. (A) Comparison of DAF between SNPs located in 5′-UTR, 3′-UTR, and coding regions; (B) Comparison of DAF between 5′-UTR SNPs within and outside of TIS; (C) Comparison of DAF between 5′-UTR SNPs located within and outside of splice site (SS); (D) Comparison of DAF between coding SNPs located within and outside of SS.

Fig. 6.—

Relationship between SNP age and DAF. (A) DAF increases with SNP age. Outliers are aggregated at the top left (red points with age < 300 kiloyears and DAF > 0.8) and bottom right (blue points with age > 700 kiloyears and DAF < 0.2) corners. SNPs located in the top right (green points with age > 700 kiloyears and DAF > 0.8) and bottom left corners (orange points with age < 300 kiloyears and DAF < 0.2) are highlighted. (B) Comparison of PhyloP score among SNPs located in the four corners in panel A. (C) Comparison of GERP score among SNPs located in the four corners in panel A. (D) Comparison of CADD score among SNPs located in the four corners in panel A.