how to determine the hybrid species in the result of exhaustive search

[bettermaan]

Dear professor,

Thanks for the convinient software !
I recently applied HyDe to detect hybridization events among 11 goups. I performed analysis for all combinations of triplets , which revealed significant signals of hybridization in 186 out of 660 combinations.
However, I found some confusing phenomenon in my data. I read many papers that involved HyDe, but I have not found the answers. I hope you can help me.

question 1: both of two below triplets are significant in my results, how can I determine whether B_eca1 or B_eca3 are hybrid.
P1 Hybrid P2 Zscore Pvalue Gamma
B_eca3 B_eca1 B_roc 27.65214517 0 0.415418127
B_eca1 B_eca3 B_eca4 22.36050636 0 0.51466702

qeestion 2: below triplets are all significant in my results. If B_eca3 is the hybrid lineage, how can I determine which group is the parental lineages of B_eca3.
P1 Hybrid P2 Zscore Pvalue Gamma
B_kan B_eca3 A_eca2 20.64633255 0 0.477761985
B_eca2 B_eca3 B_eca4 13.20803585 0 0.504983026
B_eca2 B_eca3 A_eca2 13.63157792 0 0.805693859
B_eca1 B_eca3 B_eca4 22.36050636 0 0.51466702
B_eca1 B_eca3 A_eca2 21.65425722 0 0.749721073

I guess I have to combine the HyDe result with other analysis to understand the whole story. But I wondered whether and how should I determine the hybrid species based on HyDe result.
Best,
Liu

[pblischak]

Hi Liu,

Based on your results, it looks like there may be some evidence for hybridization among the ancestral lineages of the taxa you are testing. I think that is why you are getting strong evidence for hybridization for the same taxon between different triples.

Depending on what type of data you have, I would suggest running something like SNaQ or PhyNEST, which will estimate an actual phylogenetic network that may be able to infer these ancestral dynamics better.