It depends a lot on whether you are interested in ancient recombination events or more recent ones. In the lentiviruses, for example, it is easy to study recombination events that have taken place in the past 20 to 50 years, but much more difficult to study events that took place 10,000 years ago or more. Even within the HIV-1 M group which shared a common ancestor roughly 100 years ago, events that took place very early in the pandemic are problematic because the lineages had not diversified far from the common ancestor.

So, if your species A and species B are 60% identical to each other, but isolates within each of these species are typically 95% identical to each other, it is simple to spot an A/B recombinant that recombined recently, but difficult to spot an A/B hybrid that formed long ago when A first split from B, and in fact it is most likely that any such ancient recombinant would have been classified as "species C".

If you have a "species C" and you are wondering if maybe it is more "A-like" in some regions of the genome, and more "B-like" in other regions, software like SimPlot and RDP are useful in answering that, and judging whether this is more likely to be due to recombination than to alternative selection pressures on various gene regions.

Brian Thomas Foley

Thank you so much for your helpful answer!

I have a "species C" and tries to find some sequence region which is more like "species A" and "species B".

I analysed my data set using RDP4, but there are too much "positive detection" which seems to be false.

As you said, the detection of ancient recombination seems difficult to find.