In dinaIssakova/rgenesconverged: A Package for Estimating Probability of Convergent Evolution from Genomic Data

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(rgenesconverged)

Environmental challenges impose on species an unrelenting pressure to adapt, and nowhere is this more evident than in the study of convergent evolution. Contrary to traditional assumptions of divergence, convergent evolution is the evolution of similar phenotypes in response to these challenges by species from independent evolutionary lineages. Often we want to consider `microphenotypes', or phenotypes occuring at the protein level. In such a case, two similar proteins or protein motifs can arise independently in separate lineages and have a similar genetic code without having come from a common ancestor, rendering the traditional test for sequence dissimilarity invalid. This is referred to as sequence convergence, contrasting with functional, mechanistic, and structural convergence.

rgenesconverged provides the ability to, in R:

• use Zhang and Kumar[1]'s traditional model for sequence convergence
• Explore potentially convergent data
• Obtain values of statistical significance for this data
• Plot the results

This document introduces you to dplyr's basic set of tools through the R environment. We also provide a graphical API through RShiny with a currently more limited selection of functions.

Example: Primate Data

We can explore the functionality of rgenesconverged using an example dataset of one gene across the primate evolutionary tree. This data comes from the package ape. Paradis E. & Schliep K. 2018. ape 5.0: an environment for modern phylogenetics and evolutionary analyses in R. Bioinformatics 35: 526-528.

Note that to use a dataset using the RShiny API, only the alignment is needed, and in FASTA format. Here the data is already formatted as a phylogenetic tree and a phydat alignment object. These files come pre-loaded in rgenesconverged.

print(tree)
print(primates)

Suppose we are interested in position 1, and determining whether or not it is an instance of convergence.

Let us first test this using Zhang and Kumar's [1] model.

areCondSatisfied(tree, primates, "Human", "Chimp", 2, type="abs")

We get false: this means that the residues are either equal, or one is equal to the (reconstructed by parsimony methods) ancestral state.

library(Biostrings)
data(BLOSUM62)
areCondSatisfied(tree, primates, "Human", "Chimp", pos=2, type="score", threshold=0, BLOSUM62)

What if we increase the threshold?

areCondSatisfied(tree, primates, "Human", "Chimp", pos=2, type="score", threshold=10, BLOSUM62)

Suppose we like the threshold where it is, at 0. How likely is this specific site configuration?

probOfSiteConfig(tree, primates, "Human", "Chimp", pos=2)

How likely is it that we got a score above our desgined threshold? (In other words, getting 1 match in a considered sequence of this length? Suppose, for runtime, that the sequence length is 2. )

probOfNSitesByChance(tree, primates, "Human", "Chimp", n=1, m=2, t=0, type="score")

References

[1] J. Zhang and S. Kumar, “Detection of convergent and parallel evolution at the amino acidsequence level,” Mol. Biol. Evol., vol. 14, no. 5, pp. 527–536, 1997.

dinaIssakova/rgenesconverged documentation built on Jan. 7, 2020, 4:25 p.m.