Why should we remove recombinant regions when we construct phylogenetic tree?

Question

[grandlee]

Hi everyone, I'm studying population genetics using bacteria genome.

I'm trying to construct phylogenetic tree using whole genome of bacteria. Generally, recombinant regions should be removed before constructing tree because recombinant regions have bad influences on tree. But I can't understand why need. I think the recombinant regions are one of the features of the genome, so removing that regions reduces the information of genome. To elucidate population structure using for example STRUCTURE or PCA analysis, don't we need to remove that region?

I don't have enough knowledge about this field, so I'm sorry if I've said something wrong.

Regards,

Answer 1

[liorglic]

I'm no expert on bacterial phylogeny, but I'd say that recombinant regions, which are the result of horizontal gene transfer (HGT) events, break the basic assumption of phylogenetic tree reconstruction methods. In other words, describing population evolution as a tree simply does not account for HGT, and therefore including such regions is likely to lead to erroneous inference. I believe there are other methods and tools to analyze such scenarios, e.g. phylogenetic networks. I don't know about the usage of recombinant regions in population structure and PCA analyses, but I'd say that it is likely that such regions do not resemble the overall evolution of the lineages, and should therefore be treated separately.

[grandlee]

Hi liorglic, Thank you for your reply.

I totally understood.