Measuring Accelerated Rates of Insertions and Deletions Independent of Rates of Nucleotide Substitution

Abstract

Evolutionary constraint for insertions and deletions (indels) is not necessarily equal to constraint for nucleotide substitutions for any given region of a genome. Knowing the variation in indel-specific evolutionary rates across the sequence will aid our understanding of evolutionary constraints on indels, and help us infer how indels have contributed to the evolution of the sequence. However, unlike for nucleotide substitutions, there has been no phylogenetic method that can statistically infer significantly different rates of indels across the sequence space independent of substitution rates. Here, we have developed a software that will find sites with accelerated evolutionary rates specific to indels, by introducing a scaling parameter that only applies to the indel rates and not to the nucleotide substitution rates. Using the software, we show that we can find regions of accelerated rates of indels in the protein alignments of primate genomes. We also confirm that the sites that have high rates of indels are different from the sites that have high rates of nucleotide substitutions within the protein sequences. By identifying regions with accelerated rates of indels independent of nucleotide substitutions, we will be able to better understand the impact of indel mutations on protein sequence evolution.Electronic supplementary materialThe online version of this article (doi:10.1007/s00239-016-9761-9) contains supplementary material, which is available to authorized users.

Inferring Rates and Length-Distributions of Indels Using Approximate Bayesian Computation

“…Finally, studies that aimed to elucidate how indel dynamics vary among genes and taxa rely on ad hoc methodologies for indel rate inference (e.g., Chang and Benner 2004 ). Of note, Navarro Leija et al (2016) recently explored variation in indel rate independent of the substitution rate across a given MSA. They account for this variation by presenting a stochastic model, where the gap character “-“ is added to the DNA alphabet as a fifth character.…”