Skip to content

Estimation of evolutionary networks from DNA sequences, assuming independent sites

License

Notifications You must be signed in to change notification settings

JuliaPhylo/PhyLiNC.jl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PhyLiNC

Stable Dev Build Status Coverage Code Style: Blue ColPrac: Contributor's Guide on Collaborative Practices for Community Packages Aqua

PhyLiNC is a Julia package implementing a method to estimate a phylogenetic network from DNA sequences, assuming that sites evolve independently. It does not account for incomplete lineage sorting, but accounts for possible hybridization, gene flow or recombination. The acronym stands for "phylogenetic Likelihood Network from Concatenated data".

A phylogenetic network is a graphical representation of the evolutionary history of a group of individuals (or populations, or species), in which each node represents an ancestor and either a splitting event (e.g. population split, or speciation) or a merging event (e.g. migration, gene flow, hybridization).

For details on the method, identifiability results, and semidirected NNI (sNNI) moves to move from one network to another, see Chapters 2-4 in the dissertation by Allen-Savietta (2020), which was joint work with Cécile Ané.

Note

Much of this package was formerly part of PhyloNetworks v0.16.4 (and prior). PhyloNetworks v0.17 will be stripped of PhyLiNC functions. The semidirected NNI (sNNI) moves described in Chapter 4 will remain in PhyloNetworks v0.17, as they are of general interest.

Warning

The PhyLiNC method has not been thoroughly tested: consider using NetRAX instead which was developed independently around the same time as PhyLiNC but has been through much more testing.

Below are the main differences between PhyLiNC and NetRAX.

  • NetRAX is more flexible in that it can consider 1 or more site per block, whereas PhyLiNC has the constraint that 1 site = 1 block.
  • When searching the network space, NetRAX may consider candidate networks that are non-distinguishable from the current network, which may slow it down (e.g. due to the presence vs. shrinkage of a 3-cycle; or rooted networks that only differ by their root position, identical when semidirected). PhyLiNC avoids comparing indistinguishable networks, based on theory about identifiability (see Chapter 3 in Dr. Allen-Savietta's dissertation).

Citing

See CITATION.bib for the relevant reference(s).

About

Estimation of evolutionary networks from DNA sequences, assuming independent sites

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages