Hacker News Reader

Comment by gewa | Hacker News Reader

I don't think the article addresses the point right. Yes, there is only one strain of Sars-CoV-2 as the definition of a strain is a strict scientific term. Nevertheless, there are different clades [1] of Sars-CoV-2, which are characterised as organisms with a common ancestor [2]. This doesn't has to imply a changed property, like infectivity or mortality, but it implies common heritage which is of course of interest. You can see it nicely in the the phylogenetic tree of the sequenced genomes [3].

[1] https://onlinelibrary.wiley.com/doi/abs/10.1002/jmv.25902

[2] https://en.wikipedia.org/wiki/Clade

[3] https://nextstrain.org/ncov/global

View on HN

Indeed, and you can see on nexstrain (The green RNA region labeled S) that there has been a very prolific functional mutation to the Spike Protein RNA. It is possible that the rapid growth of that strain (in Europe and NY) is due to either a founders effect or higher transmissibility (researchers do not see higher mortality, but may see higher viral load). There are possibilities that it changes the effectiveness of immune response, but it’s all under active research. The outcomes are very significant and unknown. Making broad fairly pedantic statements as the OP has, is unhelpful to understanding.

https://blogs.sciencemag.org/pipeline/archives/2020/05/07/mu...

Nitpick, but viruses are not organisms and so they don't have clades. That's why the paper you linked used the term "viral clades". Like how quasispecies is used rather than species for viruses.

I respectfully disagree. Viruses contain RNA, have reproduction dependent on DNA, and are subject to evolution, which is why they're still categorized in their own tree of "life" with kingdoms, phyla, families, genuses, species, strains, etc. There's no reason why you can't categorize a virus under a clade.

Someone reading this might be wondering how speciation can be measured with viruses, since viruses don't reproduce sexually. Well, we don't test whether different species of animals can reproduce with each other. There's no way to conduct such experiments at any meaningful scale. The vast majority of defined species are educated guesses, and in reality has very little to do with whether two groups of organisms can reproduce with each other. In a lot of cases, species are simply determined by the appearance or behavior of one group of organisms over another group of closely related organisms. The term "species" didn't even originally have anything specifically to do with reproduction; the word effectively meant "looks like", which is why the words "species" and "spectacle" share the first 4 letters.

Hold up are we arguing over semantics?

Well you don’t seem to be arguing over syntax.

> Well, we don't test whether different species of animals can reproduce with each other.

"They've been in the room an hour, sir. Nothing's happening."

"Hmmm... Ok, looks like polar bears and flamingoes are different species. Mark that down. Next on the list... swap out the flamingo for the barnacle. Be sure to keep watching closely."

It matters to me.

> Someone reading this might be wondering how speciation can be measured with viruses, since viruses don't reproduce sexually.

Viruses clearly reproduce with humans. By a strict "reproduction / species" definition, Viruses are human.

Throw in retroviruses and it gets even weirder.

That's not the strict definition. You are using an extremely loose definition. You are missing two key aspects. First, this defintion of species only applies to sexual organisms, and it's about sexual reproduction. Not just any kind of reproduction. Second, the offspring of said sexual reproduction itself must be fertile. So no, in no way can you ever say viruses are human with this definition.

They are definitely borderline; categories tend to fail to capture the continuous. But it is general scientific consensus that viruses do not meet all the criteria for life. They meet some, but not all. And no, they are not still categorized in their own tree of "life". We don't use kingdoms, phyla, etc. anymore. Your argument for viruses being life is clearly out of touch with modern biology, and since clades are by definitions groupings of organisms, and organisms by defintion are living things, no, you cannot say that "There's no reason why you can't categorize a virus under a clade". You have to either change definitions or show that viruses meet all criteria. But then if they did they wouldn't be viruses.

The species definition you brought up does not apply to asexual organisms, for them it's a matter of genomic analysis. As for sexual organisms, it is still a good definition because it is stricter and more "correct" than genomic analysis, although bioinformatics has been getting pretty good with the addition of more sophisticated statistical techniques that go past simple DNA-DNA hybridization.

Do they not use OTUs like for bacteria? (operational taxonomic units)

Yes viruses are not alive and speaking strictly there are no clades or specieses. But I stopped beeing to strict with terminology like this on HN. Most people are not from this field and don't have broad prior knowledge and so it would just complicate and confuse unnecessarily.

This is all correct, but the importance here is to combat the misinformed titles everywhere talking about multiple different "strains" going around, as well as that one damn Chinese paper (now retracted) claiming two different strains with different mortality rates.

Put more simply, two samples of virus can be of the same "strain" even if they aren't _exactly_ the same sequence. This is because replication can sometimes make errors on segments of the genome that do not impact functionality - moreover these same irrelevant loop segments are not error-corrected either, so mutations are common - despite not modifying the properties of the virus. These irrelevant mutations are actually very useful for tracking the virus spread.

A virus mutation becomes a different "strain" when the change is meaningful for the viruses interaction with the host or the environment.

SARS-COV2-2 is a type of virus that does error correction (unlike flu), so the rates of mutation are extremely low. This is good news because if means a vaccine will work, and we won't need a yearly one - like the flu.

For sure. But I think the point the @gewa trying to make is that yes in the strict academic/technical sense it is correct, however in every day English people probably use "strain" to mean Clad. But not everyone is a evolutionary biologist of some sort.

For sure. But the point of the article is that to non-expert readers, the "every day English" usage people are thinking of is misinforming them!

You are literally nitpicking the wrong people here. They're trying (well, I thought, though clearly not well enough to satisfy HN pedants) to explain the distinction in a way that conveys the fact that there are no known distinct outbreaks with different medical properties. How that's for precision?

> This is good news because if means a vaccine will work, and we won't need a yearly one - like the flu.

That's way ahead of the facts. It means we're more likely to be able to develop a vaccine that works and it's more likely we won't need frequent vaccination to be effective.

Vaccine success involves a certain amount of finger crossing, because it's a natural system rather than an engineered one that we're trying to tamper with. If you've ever boggled at a large spaghetti program, the human immune system makes that look like two dozen lines of clearly documented Java by comparison. None of it has to make sense because Mother Nature doesn't care why she only does results, and in the most brutal way possible.

This is why we've got a bunch of different vaccine programmes in different centres. Some of them might work, hopefully at least one does, and hopefully it produces an immunity that lasts a useful amount of time, has minimal side effects and is cheap to produce in bulk. But there aren't any promises without us having way more advanced biotechnology than exists anywhere today.

> A virus mutation becomes a different "strain" when the change is meaningful for the viruses interaction with the host or the environment.

Of course, that also means that detecting different strains is dependent on understanding those interactions and where they differ. If the post-infection cluster of pediatric multisystem inflammatory syndrome in NYC isn't just NYC being better at identifying that effect, it could well be a sign of a different strain with meaningfully different interactions with hosts. (It could still be a variety of other things, too.)

> If the post-infection cluster of pediatric multisystem inflammatory syndrome in NYC isn't just NYC being better at identifying that effect,

Doctors in the UK were seeing this throughout April, and issued an alert in late April: https://twitter.com/PICSociety/status/1254508725227982848

Different groups of doctors around the world are noticing different aspects of the virus. This is normal with a new virus, and does not indicate that there are multiple strains.

I know you didn't assert it, but important to keep in mind it doesn't show the contrary either - ie nor does it show it's not [now] multiple strains.

I hadn't seen the UK alert on that, but to be clear my point wasn't that that particular this was an indicator of a different strain, but that something like that, if it turned out to be isolated which wouldn't always be immediately obvious, could be such an indicator that would only be recognizable as such in retrospect.

And apparently that mutation already happened

https://www.santafenewmexican.com/news/coronavirus/lanl-find...

> SARS-COV2-2 is a type of virus that does error correction

What is the biological process that does this? I'd love to learn more.

Totally not a molecular biologist, virologist, etc.

Influenza is smaller about 13K nucleotides vs 30K for corona, it mutates about 4 times faster. It can maintain its larger genome basically because it has better tech. (It seems to better integrate with what it can find in host cells.)

"We demonstrate here, at the molecular level, that CoVs have indeed acquired an enzyme able to enhance the overall fidelity, and that this event might have directly promoted the jump in size of CoV genomes"

"Well before the discovery of nsp14-ExoN in the large Nidovirales genomes (22), this observation led to Drake’s visionary proposition that “RNA viruses would have to acquire several host genes and adapt them to RNA substrates to achieve a major reduction in spontaneous mutation rate. The result would be a substantial increase in genome size” "

https://www.pnas.org/content/115/2/E162

https://www.ncbi.nlm.nih.gov/genomes/GenomesGroup.cgi?taxid=...

https://www.ncbi.nlm.nih.gov/nuccore/1798174254

"Specifically, SARS-CoV-2 seems to have a mutation rate of less than 25 mutations per year, whereas the seasonal flu has a mutation rate of almost 50 mutations per year."

https://www.livescience.com/coronavirus-mutation-rate.html

>This doesn't has to imply a changed property, like infectivity or mortality

What about the reports of different strains that focus on having found them having different properties (such as "much more infective")? Are they bogus?

That report was from a Chinese hospital (then repeated millions of times by the media) - they were never confirmed with any sequencing, and the paper was never published due to lack of evidence.

This recent preprint certainly did not come from a Chinese hospital. The authors are listed as being affiliated with the Los Alamos National Laboratory, Duke Human Vaccine Institute, and a few Sheffield institutions.

[1] https://www.biorxiv.org/content/10.1101/2020.04.29.069054v1....

GP was referring to the “L” and “S” strain theory I believe, which was debunked. The paper you linked is new, and the “more transmissible” property they discuss may just be the “founder effect” at work. It needs more research.

It's difficult to know what user coldtea had in mind. He could have meant the L/S (debunked) theory, or the theory that some strains in Iceland are ineffective [1].

In any case, there are multiple credible sources that some of the mutations of the SARS-CoV-2 have a meaningful effect. In which case the claim of this article that "there's only one strain" may be technically correct (based on some obscure scientific definition), but completely misguiding as far as the general public is concerned.

[1] https://english.alarabiya.net/en/features/2020/03/25/Coronav...

> the theory that some strains in Iceland are ineffective

I don't see any support at all for that theory though. It is also known that "X show no symptoms" could mean many things, depending who says it and when, and certainly doesn't prove anything alone, only in context, compared to something else, and only when comparison is fair.

For example, from many old news, those "with no symptoms" eventually had the symptoms, it was just that the tests were positive before they developed the symptoms. Second, it is now known that the probability of "having" symptoms is very dependent on the age of the person.

>It's difficult to know what user coldtea had in mind

I didn't have a particular paper in mind that I remembered.

Just remember reading several articles in the past month about new strains. Some examples are easy to find:

https://www.abc.net.au/news/2020-05-06/new-coronavirus-covid...

https://www.cnbc.com/2020/05/05/the-coronavirus-mutated-and-...

https://www.usnews.com/news/health-news/articles/2020-05-06/...

nitpick! conspiratorial theories are debunked; scientific theories are refuted, no? :)

Well, a scientific theory could be bunk as well.

E.g. somebody claiming they invented a perpetual motion machine.

Their claims might not include any conspiracy involved, just a bogus idea of what they accomplished...

Probably bogus but hyper convenient and comforting so people keeps spreading it

TLDR; use “isolate” not “strain” to describe genomic differences in viruses

Strain implies functional differences. SARS-CoV-2, like it’s Coronavirus cousins, is quite stable. The genomic sequences found on NextStrain.org are a powerful tool to track the spread of this virus but it is wrong to assume that the difference in each “isolate” carries with it a functional difference; it does not.

The blog post comes from the same scientists that create the TWiV (This Week I’m Virology) podcast which continues to be a tremendous source of quality information for me.

>it is wrong to assume that the difference in each “isolate” carries with it a functional difference; it does not. //

Is that what people are assuming?

It's not that every mutation causes a [medically relevant] functional change, it's that a mutation could have. And, AFAIK, we don't have the ability to sequence _and_ relate sequences to differences in symptoms in the general case at present.

Maybe in 10 years someone will have developed an AI to relate RNA sequences, medical history, and functional changes in real time??

NextStrain is an amazing resource, it shows all sorts of viruses, not just Covid19. Really incredible.

Reply on news.ycombinator.com