I have 12 species from same genus. For supermatrix approach we generally use the multiple sequence alignment single copy orthologs sequences then concatenate them and remove the badly aligned columns and generate ML trees. In my case I got around 2000 single copy orthologs that I used.
To add outgroup species, do I first need to cluster the outgroup sequences with my 12 species sequences? If the reply is yes, then I wonder it may decrease the number of single copy orthologs as they are the possibly the part of core genome and I suppose probably they all not will be present in outgroup or may be removed by MCL clustering because of less similarity than others.
In all cases try first your tree after adding the outgroup , which should be carefully selected , say a sequence from the same family OR other taxonomic group , and probably for a gene near what you are studying - but not the same gene- , try to use different statistical parameters
good luck .
It will depend on what type of organism you are studying, and how far out the outgroup is, relative to the distances within your genus. I assume you are studying some eukaryotes and not bacteria, or archaebacteria, but even within the eukaryotes the distances between species of fungi, plants, protists, insects, bivalves, vertebrates etc can vary greatly.
With the viruses I study (mostly HIV-1 but also other lentiviruses and other viruses outside the retroviruses) I have found that using an outgroup is not always useful due to "long branches attract" problems. The outgroup often joins the clade I am studying somewhere on one of the branches even when we know from other data that the ingroup is a nearly perfect star phylogeny so the outgroup should join at the center of the star. This can lead to many types of misperceptions about the evolution if we assume that the outgroup is correctly joined.
If you have fossil data or other data to help give you solid evidence about the evolution, it can help you decide if your outgroup is helping or not.
I agree perfectly with Brian sometimes if you do not have specific problem being an out group also used to make a direction to the tree is better to use a mid point rooting that is a sort of average of a probable root due at a probable ancestor
- Badges
- Science topic
- Similar topics
- Biological Science
- Evolutionary Biology