Outgroup for phylogeny

Question

[Jaykumar]

Can anybody guide me on how to decide on which genome to use as outgroup for plotting rooted phylogeny? Does the outgroup has to be cited in literature?

I, not knowing anything, had used two criterias:

1. the chosen genome will not be part of ingroup, and
2. it will close enough (has some similarity) in order to draw core-genome phylogeny

For example, I used S. haemolyticus for rooting the S. aureus tree. Now, how do I varify/support/validate S. haemolyticus as outgroup in my phylogeny that what I have used is correct?

Thanking you all in advance!

[dthorbur]

Your criteria sound reasonable. Ideally, you want the outgroup to be quite similar to the ingroup to enable an accurate comparison, but not so similar that you risk placing it in the ingroup.

If your outgroup is placed within your ingroup, this could simply mean your assumptions about their evolutionary relationships is incorrect. An alternative explanation is that there isn't enough difference in the locus used for phylogenetic inference (in a gene-based phylogeny, for example).

You can also add multiple outgroups in an attempt to increase confidence, though the benefit of adding more outgroups will drop off quickly.

Without much knowledge of Staphylococcus, it sounds reasonable to use a species within the same genus to root a tree entirely made up of S. aureus.