The (2020) Variants: Not Guilty
Stunning blow to Brian's career as a synthetic virus detective/prosecutor.
G.O.F. Research
If you want to imagine the 2020 “Variants of Concern” as having been on trial on my laptop and in my head for the last two weeks, as in a dramatic movie, then it was Lambda, of all obscure creatures, whose discovery by the defense finally resulted in (apparently) clearing the defendants’ names.
To put it as briefly as possible, I had asked of the 2020 variants, are there any which originate from clades that existed after March, 2020, and display the bizarre mutation signatures which may or may not occur in chronic infections, but certainly do not occur in normal infections (as recorded on sequences throughout the calendar)? Gamma partially met the first condition; but as it was only one extra mutation, it wasn’t convincing evidence (to only match one local mutation, out of all the VOCs, could be luck).
Lambda was the star witness to clear the defendants’ names, as further detailed below.
Although this represents a spectacular blow to my case for continual releases, I am happy to have the better understanding of the virus’s mutation habits that this giant, obliterating failure (GOF) has garnished. Hopefully the reader will also find value in said outcome:
Conclusions from the Giant, Obliterating Failure to prove the 2020 VOCs were artificially created:
Although the late 2020 variants should not have been expected, they still occurred.
They should not be expected because RNA viruses do not actually mutate a lot, particularly long term, due to stabilizing and purifying selection, transmission bottlenecks, etc. See SVAH pt. 1.
And they should not be expected because they possess a lot of uncommon mutations without many common mutations. It is as if someone is drawing royals over and over, with hardly any number cards, from a supposedly shuffled deck. By the time someone has ten royals, they should have twice as many number cards. All the VOCs defy this by having a lot of transversions and very few C>U mutations or silent mutations, as if they were designed, rather than evolved. The VOCs accumulate overall coding mutations at 5 or more times the per-silent-mutation rate of the flu spike protein alone, and even exceed what I have been able to make out as the dN/dS ratio for mice serial passage (though this is difficult since most of the studies don’t report synonymous mutations).
Chronic infections may or may not be responsible. There are still some inconsistencies here; but if the VOCs are merely the outliers in the chronic infection mutant bucket (those without any “regular” C>U and silent mutations), then maybe there is still a link. In a way, it does not matter — suddenly accumulating multiple, coding transversion mutations with almost no “cost of doing business” c>u / silent mutations is just something the virus, in 2020, “could do.”
Essentially, between 2020-2021, how the virus mutates in normal transmission isn’t the same as how the virus evolves. The former is akin to drift with negative (stabilizing or purifying) selection against change; the latter is a sudden burst of change (far outstripping what we see in individual flu genes), not even focused on adaptive immune evasion (there is typically just one mutation in the Receptor Binding Domain before the variant takes off), with ambiguous fitness implications.
The “parachute” effect is the result.
This is a fact that must be accepted, prima facie, regardless of whether explanations besides a lab origin are satisfactory.
Large gaps between currently active clades (those that drive collected sequences on a day-to-day basis) and the VOCs remains consistent with “chronic infections” (in that those normally do seem to operate as a “chamber” of evolution over several months, as opposed to involving reinfections from transmitting virus) but also with a generic, mysterious force which makes it so that no matter how much we sequence, we do not catch the genomes that have future “staying power,” only those that are somehow already doomed.
This is, in effect, still a form of “Gain of Purity.” The purity in question is below-trend levels of accumulating C>U mutations. Since RNA viruses sustain genetic stability via purifying selection, perhaps this isn’t remarkable. The “simultaneous, parallel” accumulation of transversion mutations is “simply” a reflection of most C>U mutants dying off as the variant develops. Hence how the virus “swims upstream.”
So we’re simply in “weird dancing graphs from Akira” territory with SARS-CoV-2 from mid 2020 to November 2021. Again, just something we have to accept. Regardless of whether imagining a coterie of “immunocompromised Tetsuos” makes this fact easier to swallow or not. From henceforth, this post will refer to the mutation dynamics that lead to the VOCs as “Being Akira Lights.”
(I’m still a bit sore on the immunocompromised mutation signatures part, if you haven’t noticed, and would like to retain ambivalence about whether chronic infections “explain” anything about the variants being extravagantly poor in synonymous and C>U mutations. SARS-CoV-2 is not the virus that “invented” chronic infections, so this doesn’t give it a blanket pass to mutate in ways we don’t see in other viruses.)
Of course, all of this stops after BA.2 and BA.5; which after release have behaved as sustained “drifters” (with unremarkable positive selection on the Receptor Binding Domain, suggesting natural immune escape but without any corresponding reinfection potential). Oh, no problem! Just change the rules of reality as you go, SARS-CoV-2. It’s fine.
Why Lambda Matters for a Lab VOC Origin
My timeline of SARS-CoV-2 hopefully makes clear how bizarre this “chamber of mutation” effect truly is; but also why Lambda (and Mu, not pictured) so starkly dissolves what is otherwise a seemingly impossible consistency of “reverting to B.1/B.1.1 backbones”:
This timeline was created to impress the jury / reader with the fact that we were retroactively able to construct so much of the development of the virus in the months before the first human sequence, Wuhan-Hu-1. The “footprints” that led from the October release to Wuhan-Hu-1 did not disappear. And yet we still seem to have no footprints for Alpha, and the footprints for the Omicrons, such as they are, describe two virus families somehow traveling in different secret evolutionary landscapes only to emerge in the same place at the same time for no reason (see “Omicron Origins”).
At last, what I hoped to emphasize is the negative space. If evolving in secret, Being Akira Lights style, is just something the virus “can do,” then why don’t any of the versions besides B.1 and B.1.1 do it (I wanted to pose)? One aspect of RNA evolution to keep in mind here is clonal competition. Sure, we can imagine that VOCs happen “because evolution,” or “because fitness” (debatable), both ways of saying that they happen because otherwise the virus does not survive. But the clades after March, 2020 all should want to survive, too — they should have just as much motive, and abundantly more opportunity, to achieve Being Akira Lights, as the very earliest versions that propagated. If Being Akira Lights is to be accepted as this magical ability that doesn’t result from lab editing.
Hence why finding that Gamma and Lambda bear the signatures for their local, spring 2020 clades (even if those “signatures” are only one and four mutations, respectively) verifies Being Akira Lights for multiple extant versions of the virus, not just the earliest two on the B.1 branch.
This leaves only the following possibilities:
Being Akira Lights is a magical ability that doesn’t result from lab editing. It is effectively some sort of autonomous “Gain of Purity” in that most circulating C>U and synonymous mutations “disappear;” and existing clades are effectively doomed at all given moments. Until after the Omicrons when suddenly there is no more Being Akira Lights.
Being Akira Lights is due to lab editing, but these labs are distributed and, in the South Americas, use local spring, 2020 templates (whereas in the UK and South Africa, B.1.1 and B.1 stock are used; and Delta remains a possible natural variant given that it arises in such a high-transmission, low-sampling area).
Being Akira Lights is due to lab editing, but the lab is centralized; and yet an artificial local mutation signature of one to four nucleotides is added for the South American releases.
Everything is actually somehow because of the Warp Speed trials, from August to December. Again, Lambda (from Peru) seems to refute this one, as well as the obvious exception of no VOCs arising in the US.
Only the first bullet seems plausible; the cringey weakness of the second bullets primarily originates from Lambda and Mu (Gamma introduces other problems as well, however).
The fourth bullet remains interesting. If the widespread release of the Covid vaccines did not, in fact, accelerate the development of the VOCs (all of them likely emerge before the wide release), then it would not be via simple “breakthrough” infection that they are created. Only something about the trials; not merely the injections themselves. But Peru and Lambda remain a problem.
Lasting Controversy
At this point, as the (detective, prosecution, defense, jury and) judge telling the 2020 VOCs they are free to go, I do not expect the reader to be convinced in either direction.
As I have emphasized throughout this report, the VOCs are still manifestations of “Gain of Purity, Differently Achieved.”
Consider the implications: SARS-CoV-2 is a virus. People transmit it to each other, and it slowly accumulates mutations (with most of them being one-off, dead-ends even in the medium term). And yet all of these actual infection lineages contribute nothing to the survival of the virus long-term.
Operatively (I have expended my budget for “effectively”), the virus behaves like a multicellular organism, with most, as in almost all, “cells” (infecting virions) having a dead-end, somatic copy of the genome — whereas future generations derive from “locked-vault” germ cell copies of the genes.
But where do these “locked-vault” copies reside?
Tony Honk’s Pro Skater 3: The Winds of Change
Of course, everything about RNA virus evolution that was reviewed in “Should Variants Actually Happen?” provides glimpses of potential answers to these mysteries. If there exists a “perfect move set,” then it doesn’t matter how many defective “somatic” copies of the virus are shot out into the universe; the perfect one will survive long term. The “locked vault” is us — the virus’s host biology, which prunes away the millions of mistakes to find the best version of the virus.
But I did not write SVAH attempting to offer what, in the current understanding of RNA evolution, is the best theory for how viruses change, but rather how they stay the same.1 But maybe the former is what I have accidentally done.
Virus Evolution: Totes Solved, Everybody
There are still myriad problems with Being Akira Lights as a model of how SARS-CoV-2 survived through its first years. Again; Being Akira Lights suddenly stops after BA.2 and BA.5 (though maybe it will happen again?). And neither virus is operating in Being Akira Lights mutation mode after the split from BA.0 (with BA.1 continuing to evolve aberrantly).
So, somewhere along the line, Being Akira Lights stops being a valid strategy for long-term survival, if it ever was (after all, all the VOCs besides Delta quickly died out). We are in the era of normal mutational drift (and somehow, BA.2/5 were already in that era before they left any footprints).
The goal of the SVAH series was to introduce skepticism regarding whether the arrival of variants with not just rapid, but also unnatural mutation signatures is something we should have really expected, or if it is a sign of human tinkering.
I doubt Lambda is enough to erase the same.
Also Alpha is still extra weird (read: suspicious) for arising in a place with high sampling.
The Omicrons: Wrongful Conviction?
At the same time, in accordance with the observations about BA.2/5 in the previous section, certain questions arise about what the extraordinary mutation path traversed by the Omicrons before their arrival actually means.
One could, for example, posit that while BA.0 was in the mysterious “chamber of mutation,” one version emerged into clandestine human transmission in late 2020, and this became BA.2 and BA.5. Meanwhile, BA.0 stayed in the “chamber” another year, only for BA.1 to emerge as the result. To revisit Exhibit 1:
If, however, some sort of link exists between the AstraZeneca vector trial and the 2020 VOCs, then perhaps the Omicrons are merely a “late bloomer” in this respect.
If you derived masochistic pleasure from this gigantic leap into scientific confusion and mystery, please drop a few coins in your uncertainty-barista’s tip jar.
See also:
Simmonds, P. Aiewsakun, P. Katzourakis, A. (2019.) “Prisoners of war - host adaptation and its constraints on virus evolution.” Nat Rev Microbiol. 2019 May;17(5):321-328.
Hello and many thanks for your article.
The variants of concern were all the minds of those who made it all up. All faked and imagined. I did some imagining myself. I am not PC.
https://alphaandomegacloud.wordpress.com/2021/12/02/various-variants-covid-19/
https://alphaandomegacloud.wordpress.com/2022/05/11/100-up-and-stupid-20-variants/
An early 2020 paper on Antibody-Dependent Enhancement. My brain hurts just reading the abstract, but it seems interesting:
Molecular Mechanism for Antibody-Dependent Enhancement of Coronavirus Entry.
https://europepmc.org/article/MED/31826992