What the purpose of compare gene tree and species topologies?

Question

[lkj666]

From my point of view, if there is differences between gene tree and species tree topologies, such as we test the RF betweent the two trees. then it means that the evolution of gene is inconsistent with speices. So the gene are exogenous by horizontal gene transfer.

But i am not sure the conclusion is correct. Can anyone offer any other advice？

Answer 1

[lieven.sterck]

Personally I would not jump to such conclusions.

Indeed, the baseline assumption is then that the gene and the organism evolve differently but that does not have to indicate horizontal gene transfer or exogenous genes.

Genes (especially if you only look at a one or a few) can have very different evolution than the whole organism. The evolution of an organism is the result of evolution of it's whole genome not of a single gene.

Eg Immuno-response genes can evolve much faster than the organism, they often also have to as they need to cope with emerging diseases and such

[lkj666]

then, How can I prove whether a gene or a gene cluster is likely to be horizontally transferred? All I can think of is by GC content, codon bias, selection pressure, or the significance of gene trees and species trees.

[lieven.sterck]

Well, all you mention there an indeed be of use but will need to be correctly interpreted. As you might have read in our answers here, we don't say it can't (or won't ) be horizontal transfer, all we said was to not jump to that conclusion.

I suggest you read up on papers that deal with this issue, I'm pretty sure there are several of those around. They will likely also indicate somewhere how they determine HGT from other phenomena.

[lkj666]

Thanks for your reply and valuable suggestion.

Answer 2

[Matthias Zepper]

Well, genes are by far not the only units of organization inside genomes.

At least among mammals, there is convincing proof that a lot of genes are shared between species in conserved synteny blocks, the longest of which is the 22Mb HOXD gene cluster. The majority of human genes fall within synteny blocks of >1Mb. The Zoonomia project is creating really cool data to retrace the evolutionary processes that shaped the genomes of today's mammalian species in great detail.

On the other hand, areas of constraint (regions of the genome that display reduced rates of evolution) or acceleration (regions of the genome that display higher rates of evolution) vary greatly in size and may affect single SNPs as well as large gene-desert regions with regulatory function (e.g. such that are relevant for enhancer-hijacking). Exemplary regions of constraint are all HOX gene clusters, whereas prime examples of acceleration are olfactory genes and the immunoglobulins.

So yes, genes and other units of the genome may exhibit different rates of evolution, but that does by no means necessarily indicate horizontal gene transfer.

[lkj666]

then, How can I prove whether a gene or a gene cluster is likely to be horizontally transferred? All I can think of is by GC content, codon bias, selection pressure, or the significance of gene trees and species trees.