Polymorphism Analysis Reveals Reduced Negative Selection and Elevated Rate of Insertions and Deletions in Intrinsically Disordered Protein Regions.
Bottom Line: We also confirm previous findings that nonframeshifting indels are much more abundant in disordered regions relative to structured regions.We find that the rate of nonframeshifting indel polymorphism in intrinsically disordered regions resembles that of noncoding DNA and pseudogenes, and that large indels segregate in disordered regions in the human population.Our survey of polymorphism confirms patterns of evolution in disordered regions inferred based on longer evolutionary comparisons.
Affiliation: Department of Cell & Systems Biology, University of Toronto, Ontario, Canada.Show MeSH
Mentions: To assess whether negative selection is relaxed in disordered regions relative tostructured regions in humans, the DFE of new mutations in these regions was estimatedusing DFE-alpha (Keightley and Eyre-Walker2007), which models demographic changes explicitly, unlikePa/Ps ratios. Thisapproach uses the SFS to infer potential fitness consequences of new point mutationsin particular regions. The SFS of SNPs segregating in 105 unrelated individuals ofthe Yoruba (YRI) population of the 1000 genomes project (1000 Genomes Project Consortium 2012) was used for thisanalysis. By comparing the SFSs of nonsynonymous SNPs with synonymous SNPs, thestrength of negative selection acting upon nonsynonymous changes can be estimated.The DFE of nonsynonymous mutations is shown in figure 4 in terms of Nes, ameasure of the efficacy of negative selection, where Neis the effective population size and s is the selective coefficient.If their product Nes is <1, mutationssegregate like neutral mutations. The DFE estimated for nonsynonymous mutationswithin all regions (fig. 4) is consistentwith previous reports for humans (Haerty andPonting 2013) and other mammals (Halligan et al. 2010). If only about 5% of disordered regions,representing short linear motifs (Nguyen Ba etal. 2012), were under negative selection, then 95% of sites would befound in the nearly neutral bin (Nes< 1). This is clearly not the case, because only 24% of nonsynonymous siteswere identified as nearly neutral. This finding is instead consistent with moderatelyweaker negative selection acting in these regions compared with structured regionswith the majority of nonsynonymous changes still being selected against. Fig. 5.
Affiliation: Department of Cell & Systems Biology, University of Toronto, Ontario, Canada.