The reticulate model of phylogeny would be more adequate for describing such complicated mechanisms as hybridization between species. The computer program to construct and visualize reticulate phylogenies is called T-Rex (Tree and Reticulogram Reconstruction).The algorithm is capable of detecting contradictory signals encompassed in a phylogenetic tree and identifying possible reticulate events that may have occurred during evolution.

T-Rex is available at the following URL: www.fas.umontreal.ca/biol/casgrain/en/labo/t-rex

I hope this could help. Good luck!!

Dr. Deli.

Thank you for your help. 

Best regards

Before doing anything more fancy, please make sure that the *BEAST analyses have run long enough for proper mixing to have occurred. All ESS values should be >200. Analyses with many OTUs and many genes may have to run for several hundred million generations, especially if you do not use a strict clock.

Yes, it is possible to use Phylonet. But note that the program uses tree topology only to place onto the phylogeny (by parsimony/likelihood) the number of hybridisation events *a priori* specified by the user. You will therefore need to run several analyses, each with a different number of hybridisations, and the program does not tell you what is the optimal number, which is a drawback. Also, coalescent stochasticity (i.e. branch length) is not considered, and with only nine genes you'll probably need to explore all aspects of the data and not just topology.

What might be helpful is to generate a (rooted) chronogram for each gene, extract coalescence times for each pair of OTUs, and use those data to estimate pairwise genome divergence times under the multispecies coalescent. Plotting that distance matrix as a splits network should enable you to see more clearly between which samples hybridisation may have happened. For details, you could check the supplementary data of https://www.researchgate.net/publication/264004279_Ancient_hybridizations_among_the_ancestral_genomes_of_bread_wheat

I hope this was helpful.

Article Ancient hybridizations among the ancestral genomes of bread wheat

Thank you very much for your help, Dr. Marcussen.

A very late response, but yes, PhyloNet allows you to detect hybridization. And the current version allows you to use gene tree branch lengths as well (I don't recommend it since estimating gene tree branch lengths is very hard) and it does not assume a priori the number of reticulations.

Best,

Luay Nakhleh