A few weeks ago there was a brief side discussion on a now-locked thread about where eukaryotes fit in the tree of life (TOL), and whether or not they are are third domain of life or actually arose from within Archaea, meaning there would just be two domains of life (1 = bacteria and 2 = archaea + eukaryotes).
This has been a reasonably contentious subject over the years, as you will see if you google anything along the lines of “2 domain or 3 domain tree of life”. Recently, however, the evidence seems to be converging on the 2 domain TOL, with the Asgard archaeota being the sister group to eukaryotes.
This was the exchange I had with @Geremy in the thread:
In my last comment, I noted that the paper Geremy cited was a preprint on BioRxiv rather than a published paper, and that it would be interesting to see how the final paper looks. Well, a few days ago it was published in Nature, and there are some significant differences compared to the preprint.
In the preprint, the abstract includes this sentence:
Phylogenetic analysis does not strongly support origin of eukaryotes from within Asgard, leaning instead towards a three-domain topology, with eukaryotes branching outside archaea.
But in the final version published in Nature, this is absent from the abstract and instead the results say:
The results of our phylogenetic analysis strongly support the two-domain topology.
The affinity of the eukaryotes with the expanded Heimdallarchaeota-Wukongarchaeota branch within Asgard (Fig.1d) appears to be the most likely topology.
So, this new paper, using the largest available dataset, matches previous support for the 2D tree and eukaryotes arising from within Asgardarchaeota.
Why the change? It seems as though 30 universal proteins were analysed in the preprint to arrive at the 3D tree, but in the published version it was found that 1 of these 30 was entirely responsible for driving that relationship, while the other 29 strongly supported the 2D topology. The 1 protein (YchF) driving the 3D topology in the analysis actually supported eukaryotes arising from within bacteria, indicating that this protein likely resulted from a HGT event from bacteria to eukaryotes.
In addition to being the latest confirmation of the 2D tree, this analysis should also serve as a example to be cautious of preprints, perhaps especially on contentious topics where we can be confident that the review process will be especially thorough.
It’s also a little reassuring for me personally as someone currently making significant revisions to a manuscript previously published on BioRxiv that even papers ending up Nature don’t always get it right the first time