The effects of these slightly deleterious mutations can be partially controlled by removing mutations segregating at low frequencies (Fay et al. 2001).
These results are generally consistent with our finding that purifying selection is removing mutations in essential genes or essential parts of genes and mutations causing extreme changes in protein length.
Because no explicit bias in the rates of gene adding and removing mutations exists in our full model, we wondered whether a bias in the fitness landscape toward neutral deletions, relative to neutral duplications could be causing the streamlining pattern.
If we knew the underlying gene genealogy of locus in a sample, and the location of mutations on the gene genealogy, we could rarefy the mutations to the new mutation rate by randomly removing mutations such that on average a fraction μ′/μ of the mutations remains, and calculate the new p from the resulting genetic variation.
Regarding mutations that affect variable names, the ClientProxyProbeT probe shows silent failures only in the case in which the variable name is removed (mutation VNRemove).
In contrast, mutations that concern timestamps present silent failures only in cases in which the concrete element is removed (mutations TSEmpty and TSRemove).
In essence, the mitochondrial bottleneck could remove mutations in mtDNA from maternal lineage.
Mitochondria-targeting TALEN (mitoTALEN) has been shown to selectively remove mutation-harboring mtDNA [ 10].
De Mita et al. (2006, 2007) have hypothesized that such purifying selection might have maintained high specificity of recognition and removed mutations deleterious to the establishment of symbiosis.
Since purifying selection removes mutations from coding regions, if the gene density is higher in the core, fewer SNPs will occur there [ 9, 20, 22].
It efficiently removes mutations which cause amino acid changes in mitochondrial RC subunits possessing crucial functions, such as CO1, CO2, CO3, CYTB (see box 1).
