Resolving multifurcating at root?

Question

[rezam]

I have some trees generated by RAxML. All of them share this common property: bifurcating in all (internal) nodes except one node which is trifurcating and has at least a child that is a leaf. For example:

(((A,B),C),(D,(E,F)),G);
(D,((A,B),C),E);

I'd like to use these trees as input to some supertree program that only accepts binary trees. So now I have some questions:

1- Are these trees considered "Unrooted Binary" or "Rooted Non-Binary"?

2- What is the reason behind the above shared property? (I am CS major, sorry if this is a stupid question)

3- What would you recommend for making them binary? (I am thinking of multi2di(phy,random=TRUE) function from ape, but I'm wondering if the answer to 2nd question would question this method)

Thank You!

[karl.stamm]

If the occasional third child is always a leaf and not another subtree, then you could delete it; and consider it instead a property of the parent. Thus the parent object has two proper subtree children and an optional (nullable) property. Then you have a binary tree with labels.