Phylogenetic species tree construction based on presence and absence of a few genes

Question

[Ritvik]

Hi,

Is it possible to construct a phylogenetic species tree based on presence and absence of certain genes? If possible, would it be evolutionarily correct?

Answer 1

[h.mon]

Yes it is possible. One simple way is constructing a character matrix for presence / absence of genes and performing a parsimony analysis.

It is correct (at least under certain assumptions), but may prove difficult. First, your genes should be well behaved, that is, genes without a lot of horizontal transmission. Second, it would be better if you were able to distinguish independent gains / losses and code them as different characters on the matrix.

[Ritvik]

Thanks for replying, ok, i will try what you have suggested. Can you suggest me some paper in which a similar strategy has been followed? The closest that i could find was this: http://www.biomedcentral.com/1471-2164/12/430/figure/F4 in which i think the authors are just showing the presence and absence of genes on a tree made from some other data.

[h.mon]

Sorry for the late reply. I found this presentation, which in fact recommends against parsimony. It seems old (newest reference is from 2004), but searching for the references you will find recent papers.

[Ritvik]

Ok and thanks for the reply, will sure check what you have suggested!