Halomonas bacteriophage vB_HmeY_H4907 discovered recently in the Mariana trench is apparently highly homologous to its host:

To our best knowledge, it is the deepest isolated siphovirus from the ocean. Its 40,452 bp linear dsDNA genome has 57.64% GC content and 55 open reading frames, and it is highly homologous to its host.

Naïvely, this seems to suggests to be an evidence in favor of a particular hypothesis of the origin of viruses - as runaway genetic elements.


  • Is it common for a viral genome to be homologous to its host (or another cellular organism)?
  • Is such homology an evidence in favor of a particular hypothesis of the origin of viruses - as runaway genetic elements?

Identification and genomic analysis of temperate Halomonas bacteriophage vB_HmeY_H4907 from the surface sediment of the Mariana Trench at a depth of 8,900 m

For laymen interpretation see, e.g., Virus from depths of the Mariana Trench uncovered or Deepest Virus Ever Detected Unearthed by Scientists in the Mariana Trench

  • $\begingroup$ Funny to see something "unearthed" in the ocean... Language can be wonderful at times. $\endgroup$
    – Jon Custer
    Commented Sep 29, 2023 at 16:38
  • $\begingroup$ Please read the paper and find out in what specific way the authors of the paper think the virus is “homologous” to the host (and indeed tell us what the host is). Then modify your question accordingly. By itself the word “homologous” is unclear. $\endgroup$
    – David
    Commented Sep 29, 2023 at 22:10
  • 2
    $\begingroup$ @David I think the second sentence in the quote, where the first half of the sentence is talking about the genome, indicates sequence homology. I'm not sure any other homology would make sense in context. Sure, I think this could be clarified, but the authors are making a relatively qualitative statement here anyways. $\endgroup$
    – Bryan Krause
    Commented Sep 29, 2023 at 22:22
  • $\begingroup$ @BryanKrause DNA, RNA, encoded protein? For how many genes? The question does not say and is therefore unclear. I have asked the poster to clarify and think that perfectly reasonable. Too many questions (and answers) are of little use here because their subject is not clearly defined. This can be because the poster assumes everyone knows what he is talking about, but often is because he is not clear about the problem. In this case going back to sort this out is an important learning experience. If the poster is not prepared to engage with us we should not engage with him. $\endgroup$
    – David
    Commented Sep 29, 2023 at 23:01

1 Answer 1


Phage similarity to host

Briefly, yes. Generally speaking phage have relatively similar compositional features (e.g. k-mer frequency, see linked paper) and also direct homology (large regions of sequence similarity) to their host genome, compared to a null model. Methods leveraging this have been used (with some success) to predict hosts.

Support for hypotheses of origin

I think that there are multiple possible mechanisms for how similarity arises:

  1. Gene conversion between phage-host
  2. accidental packaging of host genome
  3. phage integration as endovirus (possibly as a "dead" copy) and neofunctionalization.

I think that these mechanisms don't directly contradict your hypothesis. In fact, the presence of auxiliary metabolic genes in phage might argue in favor of it.

But at the end of the day, the phage and the host are just swapping DNA at a nonzero rate, and you'd expect them to be more similar than otherwise.

  • $\begingroup$ “Compositional features”? The question is not clear on what it means by homology, and your answer does not say what compositional features you are talking about. Size? An RNA genome surrounded by a protein coat? A genome that does not encode ribosomal components? $\endgroup$
    – David
    Commented Sep 29, 2023 at 22:16
  • 1
    $\begingroup$ @David "composition" is usually used to refer to nucleotide, GC, or k-mer frequencies in the context of DNA sequences. "Homology" is generally understood in genomics to refer to regions of sequence similarity. (As described in the linked paper.) I have updated the answer to reflect this. $\endgroup$ Commented Sep 29, 2023 at 22:25
  • $\begingroup$ As I implied in the original version of my request to the OP, to the man with a hammer, everything looks like a nail. Composition means composition. In the context of a genome it could mean the number and functional distribution of genes, whether they were ss, ds, RNA or DNA, their base composition etc. And your answer ignores the vast differences between viruses in the same organism: simple RNA phages like Qbeta, v. large T5. Even human herpes viruses show vastly different GC contents, and, of course they are quite different from polio and from SV40. $\endgroup$
    – David
    Commented Sep 29, 2023 at 22:49
  • $\begingroup$ @David certainly- variation exists. The question, as I understand it, is whether that variation shows any association to the host, or whether it is agnostic to the host. If you are frustrated with genomics terminology at large, and the discipline's attempt to reduce everything to sequence data, then I will happily agree that it is annoying. $\endgroup$ Commented Sep 29, 2023 at 22:53
  • $\begingroup$ I am not frustrated with genomics terminology, I resent imprecise terminology and lack of context that make statements ambiguous and scientific discussion pointless. I know about viruses, I know about relationships between viruses and hosts, viruses and viruses and genes within viruses. I have now read the paper, and have, I think, managed to fathom what the authors mean in their sloppy usage of the term "homologous". As their first language is not English they are only partly to blame for this — the journal editors should take their share. If I chose to answer this question I will explain. $\endgroup$
    – David
    Commented Sep 30, 2023 at 17:02

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .