Coronavirus variants contradict the predictions of genetic entropy proponents

Except that beneficial mutations are comparatively rare (which no serious geneticist would deny), this is not what GE asserts, not at all.

Someone organized a symposium? Well I guess that means there’s an invisible, yet to be quantified factor called “information” in his genetic entropy idea after all. Or what?

1 Like

No. GE doesn’t claim that some adaptive genetic changes cannot occur that can be selected. What it says is that while some changes at some positions can and will be positively selected by natural selection, deleterious mutations in the area not under selection are carried along with the ones under positive selection and that in the long run, this will negatively affect fitness. So none of what’s going on currently with sars-cov2 contradicts the predictions of GE proponents.

I’m letting you know. Please provide the data supporting your assertions.

We don’t misunderstand. Sanford changes the hypothesis. For a scientific hypothesis to be useful, the hypothesis itself must be stated clearly enough so that the predictions of the hypothesis are also clear. Whether Sanford states a prediction is irrelevant, except as a pseudoscientific marker.

That is a claim, not a prediction.

That is a prediction that is empirically false.

That is not a prediction but another pseudoscientific claim, with plenty of ill-defined terms that allow Sanford and you to obfuscate after the fact.

In real science, such assumptions are baked into the hypothesis and predictions are about the data, not its interpretation.

If the hypothesis doesn’t clearly predict the data independently of what Sanford says, it’s not scientific.

1 Like

How can this be tested? Suppose that 9 months from now herd immunity (via immunizations and immune response to infections) has caused Sars-Cov-2 to disappear, or nearly disappear, as a threat. The only evidence we would would be left with would be the many variants that exhibited gains in reproductive fitness, right?


Which makes GE a practically unfalsifiable idea, because now GE proponents have invented some ad-hoc excuse they can invoke every time fitness doesn’t actually appear to be declining at all. If we see fitness stay unchanged, or even go up, the GE proponent can now just baselessly rationalize that there are still lots of deleterious mutations accumulating, but they’re just being masked by a few rare beneficial mutations of large effect.

How, then, would one observationally falsify GE? If fitness increase won’t do, then how is that even possible? Does GE predict anything that could really be realistically incompatible with the data on human timescales? Do we have to wait 50.000 years, or a million, before we have enough data?


Yeah, that does indeed pretty well sum up how ignorant Sanford is about biological information – at least it does when you learn that the symposium “at Cornell University” was a bunch of creationists and IDists who hired an auditorium at Cornell’s school of hotel management to put on a meeting unaffiliated with the university.


I also have to say that the GE rationalization that “rare” beneficial mutations of “large” effect are “masking” the real fitness impact of deleterious mutations of small effect is logically incoherent, and demonstrably incompatible with Sanford’s imagined DFE of mutations.

Since Sanford basically halluscinates a DFE of mutations where beneficial mutations of significant effect are practically miraculously rare, one wonders why it is they keep happening so frequently and are of such magnitude that they appear to overwhelm the combined deleterious effects of all the deleterious mutations.

I mean, look at Sanford’s figure here again. He’s saying the combined deleterious effects of all the mutations that occur here:
not enough

Just aren’t enough to overwhelm the fitness effects of mutations that miraculously keep occurring waaaaay out here :
miraculously rare beneficials
(or much further out still, try extending that figure to include a mutation with a 5% increase in fitness)

And yet please try to consider what the probability of a mutation that far out on an exponential decline must be, considering how small Sanford has made the curve for beneficials.

This can’t be entertained by a thinking person.


Exactly. There are thousands of different viruses. Some types, especially the ones lacking genome proofreading mechanisms, have thousands or millions of variants existing in mutant clouds. Each variant may be produced by mutations, recombinations, reassortments or a combination of two or more. If we adopt a YEC timeframe, these viruses have had a few thousand years to replicate and experience the above processes, yet viruses haven’t gone extinct considering their relatively fast generation times.

Its truly an empirically false claim.


Hi Gilbert,

The question is whether Sanford’s theory of genetic entropy incorporates information theory in a coherent and quantifiable manner, right?

Whether Sanford understands what other people say about biological information is irrelevant to the issue of whether Sanford has incorporated biological information into the genetic entropy hypothesis. Would you agree or disagree?

@Rumraket 's critique of genetic entropy was not an ad hominem argument; he offered an extended analysis of its shortcomings with respect to information theory. Therefore, your pro hominem statements about Sanford don’t really have any impact, at least on my opinion, and probably on that of other readers.

I would be interested in any substantive response you might care to give to @Rumraket 's argument, though.



At this point, GE has been thoroughly discredited as a supported scientific hypothesis. The current responses we still receive from GE proponents are usually predicated on semantic wiggling rather than data-driven falsifiable hypotheses.

  • No, simulations in Mendel’s Accountant do not constitute valid scientific evidence
  • No, changing the definitions of ‘fitness,’ ‘deleterious,’ ‘beneficial,’ ‘neutral,’ ‘function,’ ‘information,’ ‘adding,’ ‘improvement,’ ‘dead,’ ‘extinction,’ or ‘degeneration’ is not compelling
  • No, claiming GE is a prediction that cannot be tested until some unspecified time in the future is not a mode for establishing scientific precepts

I think if we are going to entertain more conversations about GE, the proponents need to posit a falsifiable and currently testable hypothesis using real-world sequencing data.

At a minimum, I would like to see:

  • a clear and concise null hypothesis
  • methodological approach and data used
  • specific quantifiable expected outcomes

No need to publish a 300-page manuscript. This information can easily be conveyed in 1-2 pages.


Others have complained about this one. It is a common assertion among creationist and ID advocates that there is some principle of physics or of logic that information cannot be increased by evolutionary processes. Natural selection involves preferential increase of a subset of the population, which is a fundamental operation of Shannon’s information theory, and is also fundamental to specified information (as defined by Leslie Orgel and used by Jack Szostak, Robert Hazen, and invoked by William Dembski).

At which point, creationists and ID advocates usually say, “oh, but it’s the mutation that creates the information, all the selection does is change the frequencies of alleles that are already there”. That definition of information is in contradiction to Claude Shannon and in contradiction to specified information (and functional information). And mutation can change a base in the genome from one which has lower fitness to one which has higher fitness. We know that because you can mutate from any base to any other base at any point in the genome.

I’ve said this lots of times before. Sorry to be tiresome, but the misconception persists that there is some principle of science or of logic that natural evolutionary processes cannot increase the amount of information (or of specified information) in the genome.


GE is a slow process. Therefore, it is not at all surprising that at the beginning of an outbreak, a virus can adapt rapidly to its new environment through a serie of beneficial mutations until it reach a natural optimum, at which stage the relentless accumulation of slightly deleterious mutations will begin to slowly take effect. So no, I don’t see anything incoherent with the fact that at the beginning of an outbreak, “rare” beneficial mutations of “large” effect are “masking” the real fitness impact of deleterious mutations of small effect.

Right. Given that GE is a slow process, in the event that sars-cov2 would disappear 9 months from now, it would probably be impossible to demonstrate or observe its effect. But if sars-cov2 was to be with us for decades, then my prediction would be that we would see the effect of GE.

The ancestors of hCoV-19 or SARS-CoV-2 have been with us (in bats and other animal reservoirs, except humans) for many decades, but they haven’t gone extinct as expected under GE. Why?


What comes after the word “Therefore” in the above sentence doesn’t actually follow from the preceding statement (that “GE is a slow process”). The slowness of GE does not logically entail what should be occurring at the beginning of a new viral outbreak.

But it had not before, in whatever other species the virus came from? I suppose here we are to posit that viruses in general are just totally dormant in a frozen state? Somehow, in your idiosyncratic old-Earth view of Genetic Entropy, viruses have persisted for billions of years by periodically re-emerging from thousand-year dormancies to terrorize still-not-extinct animal populations, that mysteriously failed to evolve immunity the last time such a virus came out of dormancy. Or maybe what you think happens is a virus emerges from melting ice, infects a population for at time, reaches some local optimum, goes into slow fitness decline, then gets back into dormancy for thousands of years, then re-emerges to infect another host (or the same one, or both), re-adapts to some new local optimum, the cycle repeats where the virus reaches a new local optimum and then starts declining again, and this… means evolution is false and everything is headed to extinction. Except, a mysterious force is keeping it all going, except in bacteria, which maybe are maybe aren’t succumbing to GE because they’re simple and have large populations and low rates of mutation so here the mysterious 3rd force isn’t required.

Did I get this right?

So to begin with, the beneficial mutations of large effect aren’t all that rare at all, in fact they must start out by being rather likely so that they measurably occur, so the distribution in a new environment starts by looking something like this:
mysterious highly adaptive beneficials

But doesn’t that imply that since they begin being rather likely, that there must be a very large number of them. Otherwise, why are they likely to begin with? What is it that made the D to G mutation at position 614 in the spike protein likely to begin with, but some other substitution in a protein much later in the pandemic is very unlikely?


Hi Gilbert,

Now that I think more deeply on the issue, I believe we already have all the data we need. Consider:

Influenza and coronavirus strains have been in circulation for many millenia (at least!).

Now let’s ask: Is there any ancestral relationship between today’s strains and the strains that prevailed thousands of years ago?

If the answer is yes, let’s ask another question: How would there be any influenza or coronavirus strains in existence today if the strains that existed in 2000 BCE had to inevitably, per the hypothesis of genetic entropy, go into decline and demise?


1 Like

Its obvious HIV-1, group M disagrees after over a hundred years in existence, with insane mutation rates and very fast generation times.


The hypothesis is that these viruses can exist in a somewhat dormant state in their natural reservoirs.

Your reply poses more questions than it answers.

So genetic entropy is a strong and ineluctable force…

except it is completely absent in “natural reservoirs” for millions of generations.

  1. What are the characteristics of a natural reservoir that allow virus strains to avoid genetic entropy?
  2. If natural reservoirs exist for viruses, why can they not exist for other species?
  3. If natural reservoirs can exist for a broad range of species, how can genetic entropy make sensible and falsifiable predictions?


1 Like