Adam, Eve, and human population genetics, part 11: addressing critics—Poythress, chimpanzees, and DNA identity (continued) | The BioLogos Forum

Faith & Science Conversation
1

Note: In this series, we explore the genetic evidence that indicates humans became a separate species as a substantial population, rather than descending uniquely from an ancestral pair.

In the last two posts in this series (here and here), we have been exploring the anti-evolutionary arguments of Dr. Vern Poythress in his brief book Did Adam Exist? – specifically, his arguments alleging that that the human and chimpanzee genomes are only 70% identical. As we have seen, this line of argumentation is intended to call human – chimpanzee common ancestry into question, in part because shared ancestry is used in some methods of estimating the ancestral population size of our lineage. Since Poythress intends to argue that human ancestral population genetics cannot establish that humans descend from a population, rather than a pair, attempting to undermine confidence in common ancestry is part of the overall thrust of the book.

As we have seen, the argument in Did Adam Exist? for ~70% identity is as follows (pp. 7-8):

“If the comparison focuses only on substitutions within aligned protein-coding regions, the match is 99 percent. Indels constitute roughly a 3 percent difference in addition to the 1 percent for substitutions, leading to the figure of 96 percent offered by the NIH… But we have only begun. The 96 percent figure deals only with DNA regions for which an alignment or partially matching sequence can be found. It turns out that not all the regions of human DNA align with chimp DNA. A technical article in 2002 reported that 28 percent of the total DNA had to be excluded because of alignment problems, and that “for 7% of the chimpanzee sequences, no region with similarity could be detected in the human genome.”

Even when there is alignment, the alignment with other primate DNA may be closer than the alignment with chimp DNA: “For about 23% of our genome, we share no immediate genetic ancestry with our closest living relative, the chimpanzee. This encompasses genes and exons to the same extent as intergenic regions.” The study in question analyzed similarities with orangutan, gorilla, and rhesus monkey, and found cases where human DNA aligns better with one of them than with chimpanzees.

We have already dealt with the first section of the argument, and showed that the 2002 paper Poythress cites cannot bear the weight of his claims for it. Accordingly, we now turn to the second section of the argument – but as we shall see, it too fails to establish what Poythress claims for it.

The main thrust of the argument here seems to be that if a sizeable portion of our genome more closely matches a species other than chimpanzee, then a sizeable portion of our genome is not all that similar to the chimpanzee genome. This is an incorrect assumption, but it will take some effort to understand why.

Perhaps the best way to appreciate how this works is to look at actual DNA sequences. Shown below is a section of 90 bases of DNA from the human, chimpanzee and gorilla genomes:

There are a few things to notice here. First, all three sequences are highly identical to each other, with only a few differences. Second, the human and gorilla sequences are the closest match to each other, with only one difference present. Third, the human and chimpanzee sequences have three differences. This pattern is representative for the sequence nearby as well, outside of what can be shown in a small figure. For a stretch of DNA in this area of the genome, the human and gorilla versions match more closely (98.5% identical) than do the human and chimpanzee versions (98.2% identical). As such, this genome region is part of the 23% of our genome that does not find its closest match in chimpanzees, but rather in gorillas.

The problem for Poythress’s argument should be immediately apparent. The fact that 23% of our genome matches the gorilla genome more closely does not imply that this very same DNA does not also match the chimpanzee genome with very high identity. In fact, this portion of our genome is highly identical to chimpanzees – it’s just that it’s slightly more identical to gorillas. The issue is that the human, chimpanzee and gorilla genomes are all highly identical to each other – but that for some portions of our genome, we are just a tiny bit more identical to gorillas than to chimps.

This observation, of course, raises the question of how this pattern arises between these three genomes. The answer, as it happens, is something we have already discussed in this series: incomplete lineage sorting. Since humans, chimpanzees, and gorillas are all descendants of a common ancestral population, we expect this phenomenon to occur, since not every DNA variant in the original population (i.e the one ancestral to all three descendant species) will sort down to all three modern-day species:

So, what Poythress sees as a problem for common ancestry does not, in fact, support his claim that large swaths of the human genome are dissimilar to the chimpanzee genome. Perhaps the confusion arose from failing to understand just how similar the genomes of humans, gorillas, and chimpanzees are – and that more closely matching gorillas is not at all an indication of dissimilarity with chimpanzees. The pattern produced by incomplete lineage sorting is (ironically enough) predicted by common ancestry – shared ancestry between three species. So, far from being a problem for common ancestry, this pattern is genome-wide evidence for common ancestry.

Having dealt with the human – chimpanzee identity issue, Poythress then shifts his focus to population genetics. His intent is to argue that the evidence we have discussed in this series is unable to conclusively assert that our species descends from a population, rather than a pair. As we will see in the next post in this series, however, this claim is similarly bereft of scientific support.

Further reading on the scientific and theological issues related to Adam and Eve:

Note: this list is mostly drawn from my BioLogos colleague Ted Davis's excellent series on Evolution and Original Sin, found here.

This is a companion discussion topic for the original entry at https://biologos.org/blog/adam-eve-and-human-population-genetics-part-11-addressing-criticspoythress
4

from the article:

"The pattern produced by incomplete lineage sorting is (ironically enough) predicted by common ancestry – shared ancestry between three species. So, far from being a problem for common ancestry, this pattern is genome-wide evidence for common ancestry. "

its fit to id as well. even without the commondescent option we will find such a case. so it can predict also by id. evolution doesnt need here. and what about a case that 50% of the genes told different story?:

“Roughly 50 per cent of its genes have one evolutionary history and 50 per cent another,” Syvanen says

is this also predict by commondescent? where is the limit actually?60%?80%?

5

Hi dcscccc,

You say that ID can also account for the pattern produced by incomplete lineage sorting - how so? The folks from the Discovery Institute don’t seem to think so - they deny the existence of ILS outright. How do you see ID predicting an ILS pattern?

The second quotes pertain to rampant horizontal gene transfer between microorganisms. Primates - such as humans, chimpanzees, gorillas, and the population(s) that gave rise to them, do not experience horizontal gene transfer. So, those quotes aren’t relevant to this issue. Horizontal gene transfer can make some phylogenies challenging, but the primate tree isn’t one of them.

6

hi dr venema.

the 50% case is actually about animals(sea squirts) and not microorganisms. so the question hold water.

about the ils- id predict that we will find a lots of cases that some parts of the genome at species a will be closer to some parts at species c more then at species b for example. this is because they dont have a commondescent. very simple. the evidence fit to id and not evolution . because evolution predict hierarchy in general and id doesnt.

7

Hi dcscccc,

It seems you’re correct that the quote is about animals - my mistake, sorry. But, as far as I can see, the results Syvanen is discussing were never published. They’re merely quoted in a popular article. So, these are not verifiable results unless they are published and peer reviewed. Even so, ILS could produce such a pattern, yes.

You still haven’t explained why three species should have an ILS pattern if they are not related. Why would a designer independently create three species with this pattern when there is no need, biologically, for them to have this pattern? That’s the question. Your answer is ad hoc.

8

hi again dr venema.

if ils can produce 30% and 50% and so on- then it itself an ad hoc explanation. according to this , even if chimp was closer to rat then a human (from genetic prespective), we can still claim that chimp is closer to human then rat, and solve this by ils model.

why the designer designed animals with no hierarchy? why not actually? we find this all the time in human design. if you will take 3 different cars- you may see that some parts are closer between 2 of them and some between the rest.

9

ILS is most certainly not an ad hoc explanation - you’ll notice that ILS is only observed between species that are known to be very close relatives. In the case of humans, that’s gorillas, and to a much lesser extent orangutans. It’s also dependent on the population sizes and speciation times, as we have discussed previously in this series. We don’t see ILS with things like rats or mice and humans, because we are much too distantly related to them for ILS to be present.

So, it’s a nested, self-reinforcing pattern. Overall, we are most closely related to chimpanzees, then to gorillas, and then to orangutans at the DNA identity level. The ILS patterns we see in the human genome also match this: we have the most ILS with gorillas, and then with orangutans (though it is a tiny amount with orangs, due to our distance from them). Percent DNA identity is maintained for much longer periods; ILS is a “short” -i.e. few million years- range effect.

You know, I get that you reject evolution for theological reasons - lots of folks do. Even with that rejection, though, you’d do well to inform yourself about what we actually see when comparing genomes. I’d recommend that you read some of Todd Wood’s material, starting with his 2006 paper. Todd Wood is a YEC, but he also addresses the evidence carefully.

10

This post by Todd would also be a propos:

and here’s his 2006 paper:

http://www.creationbiology.org/content.aspx?page_id=22&club_id=201240&module_id=36954

11

hi again dr venema.

you said:

“you’ll notice that ILS is only observed between species that are known to be very close relatives.”-

about 100 milion years from now, the current chimp\gorila\human genomes will not be close at all. but the ils will stay in their genomes. so it isnt only in close relatives.

you said:

“So, it’s a nested, self-reinforcing pattern. Overall, we are most closely related to chimpanzees, then to gorillas, and then to orangutans at the DNA identity level. The ILS patterns we see in the human genome also match this: we have the most ILS with gorillas,”

they are actually the same- ils is about bases that doesnt match the standard phylogeny. so we actually expect to see this pattern if we already know what the whole genome phylogeny is.

12

@dcscccc

The reason we see a lot of ILS between Humans, Chimps and Gorillas is because a very short amount of time passed between the splitting of HCs and Gorillas and then the later splitting between humans and chimps. This time difference between the first split and the later split could have been as short as two million years. This is why we see a clear ILS signal.

This has all been explained to you many times before and yet you keep bringing this up as if it was news to you. At this point I have to believe that you are just being an obstructionist and arguing for ideological reasons even though you know the answers.

13

There is the difficulty of circular reasoning however, Ace. The ILS(incomplete lineage sorting) signal indicates a split about 2 million years ago, based on a consistent mutation rate, so the split about two million years ago is the reason for the clear ILS signal. But how do we know there was split 2 million years ago? because of the ILS signal. How do we know the ILS is a signal of anything? Because the split happened 2 million years ago. (I know 2 million years seems very short to you… but to me it is a long time.)

You still haven’t explained why three species should have an ILS pattern if they are not related. Why would a designer independently create three species with this pattern when there is no need, biologically, for them to have this pattern? That’s the question. Your answer is ad hoc.(Dennis)

Part of the question is whether this ILS, a vague term for a type of presumed process, rather than a description of actual functional dna, whether this dna has any uses that are pertinent to the different species. ILS is a descriptor it seems to me, based on comparisons with other selected species, which are selected on an assumed similarity of relationship. The word “lineage” is presumptive… that a lineage actually must necessarily exist, rather than the dna differences merely being necessary to distinguish the species based on physiology and morphology. “Incomplete” means that there are differences; “incomplete” is the difference, not similarity between species. “Sorting” refers to how one species may sometimes be more similar to species x and sometimes more similar to species y. The conclusion in this case is that there must have been a common ancestor for all three species, which allowed many similarities, and a few differences between all three. But on the whole, I find this fuzzy logic. It implies that things cannot be similar unless they are related or have common ancestry… this “law” seems to go beyond the boundaries of what we can actually demonstrate, and seems to rule out the possibility of a common designer on an a-priori basis. Also, too soon and too often something is presumed to have no biological use or need (as in the above quote), and then later discovered indeed to be very useful and purposeful. Just because we do not know the need now, does not mean that there is no need. “Vestigial” organs which are not vestigial and are very useful, are the perfect example of this fallacy.

When the differences between species are small, we would naturally expect the differences in the dna program to be small; this is regardless of ancestry or inheritance. So for you, 2 million years is short. To me, 2 my is very long. To you, 4% difference in genome is small, but to me, 30 mbp (1%) difference is huge.

14

Hi John

There seems to be some misunderstanding here. It’s not that there was a split 2 million years ago (from the present), it’s that there was an initial split between [Human Chimp common ancestor] and Gorillas and then a later split (about 2 million years later) between Humans and Chimps. This initial split might have happened about 8 - 12 million years ago.

There is no circular reasoning here because we estimate the divergence between [Humans, Chimps], [Humans, Gorillas] and [Chimps, Gorillas] independently based on things like molecular clocks.

Here are a range of published results from different papers giving the estimated divergence time between Humans and Gorillas - the most recent results indicating something between 13 and 8 million years. (scroll down to see all the papers - there are well over 20)

Here are a range of published results from different papers giving the estimated divergence time between Chimpanzees and Gorillas. As expected this is roughly the same - the most recent results indicating something between 13 and 8 million years. (scroll down to see all the papers - there are well over 20)

Here are a range of published results from different papers giving the estimated divergence time between Chimpanzees and Humans. As expected this is a little less - the most recent results indicating something between 7.5 and 6.1 million years. (scroll down to see all the papers - there are well over 20)

We can then use simple maths to subtract [HC] from [HG] to calculate that these two splits followed on quickly after each other (some estimates would say about 2 million years).

We can then independently look for how much ILS we can detect between these 3 species and we find (this should be to the surprise of creationists everywhere) that there should be about 2 million years between the two splits.

whether this dna has any uses that are pertinent to the different species

Some of this DNA will be functional and some will be non-functional but this is irrelevant to the matter at hand.

The word “lineage” is presumptive… that a lineage actually must necessarily exist, rather than the dna differences merely being necessary to distinguish the species based on physiology and morphology.

It implies that things cannot be similar unless they are related or have common ancestry… this “law” seems to go beyond the boundaries of what we can actually demonstrate, and seems to rule out the possibility of a common designer on an a-priori basis.

You only think it’s presumptive because you don’t understand the evidence. I’d encourage you to look into cases of ERVs that are shared exclusively between humans and chimps and then also humans, chimps and gorillas.

There are viruses that we know worked their way into our genomes because they leave a number of tell-tale markers and we have witnessed them inserting themselves into genes.

If the same virus of the same class and subclass with the same mutations inserted itself into an identical location in many different species leaving the same tell-tale markers of insertion in all of them then the only reasonable explanation is that this happened once in a common ancestor and that this fossil of a sequence was inherited to all subsequent species that would go on to diverge from that point.

So for you, 2 million years is short. To me, 2 my is very long. To you, 4% difference in genome is small, but to me, 30 mbp (1%) difference is huge.

That’s why we measure things like time and DNA differences relatively. 30mbp is small compared to a 3bbp genome. 2 million years is short compared to the 50 million years of primate evolution. When you throw numbers out like this, you have to consider the scales which are relevant to the topic at hand.

15

hi ace.

here is dr venema claim: “you’ll notice that ILS is only observed between species that are known to be very close relatives”

so i gave a theoretical example that ils can be found in a very far species. i dont see any problem here.

i agree with john claim that this is circular reasining. beacuse we dont know that those species actually split from each other. the molecular clock have the same problem. for example: we need another method to fit with the molacular clock. you gave the fossils. but fossil evidence change all the time. so one day the molecular clock for specific species can be 50my and another day it will be 100 my. so we cant know it is true. another problem is that we actually dont know what is junk and what isnt (like john said). so is a big problem for molecular clock.

you gave to john the erv example. but we now know that ervs are actually very functional:

so they can be the result of commondesign.

16

The problem here is that you didn’t understand him. Instead of trying to understand Dennis you made up your own interpretation of what he was saying and made up some irrelevant example.

ILS cannot be observed between two species - by definition it is a phenomenon that can only be observed between three or more species. Obviously what Dennis meant then wasn’t that the split happened that in recent history but rather that the split between two of the species happened close in time to the earlier split between all three species.

i agree with john claim that this is circular reasining.

That’s because you don’t understand it.

Of course we do.

for example: we need another method to fit with the molacular clock. you gave the fossils. but fossil evidence change all the time. so one day the molecular clock for specific species can be 50my and another day it will be 100 my.

This is complete and utter nonsense. We can measure molecular clocks between humans today by looking at how many mutations children have compared to their parents. We can do the same for Chimpanzees. Once again, this has been pointed out to you in the past so I have to think that you are only acting ignorant for the sake of appearing to have something to say in response.

Here are two methods for measuring the mutation rate. Please read these and try to remember them for the next time you bring this up in conversation:

you gave to john the erv example. but we now know that ervs are actually very functional:

It is irrelevant that some of them have gained functions. It is also entirely expected and unsurprising. The only relevant fact is that we know for sure that those ERVs weren’t originally there. They have distinct biochemical markers showing that they were once RNA retroviruses that inserted themselves into these positions.

Specifically they almost all have identical LTRs which are formed in the reverse transcription process from RNA → DNA and they almost all have identical TSRs (target site repeats) - this is when a bit of the target site is duplicated when the ERV integrates itself into that p[osition.

17

ace,

first- you claim that “The only relevant fact is that we know for sure that those ERVs weren’t originally there”

not at all. we actually have 4 evidences that those ervs are the result of design and have been in the genome from the beginning:

1)we know that a lots of them are functional
2)retrovirus is a parasite. so its origin need to be from the host genome and not the opossite
3)we have evidence that the virus itself “infected” by the host
4)we find erv(s?) that contradict the phylogenetic tree

those 4 evidences can disprove any claim about ervs.

you said: “We can measure molecular clocks between humans today by looking at how many mutations children have compared to their parents”

meybe its because of my english. i actually talk about something else.i mean that we need to find a correlation between molecular clock and other method that fit to the clock. for example: if the molecular clock of a flower species give us a result of 50 my- then the fossil record of the flower need to fit with this data. but fossil record change all the time, so we dont have any evidence that molecular clock can be credible. i also gave the marbled lungfish example.

18

That’s not evidence they weren’t originally there. We expect in biology that sometimes genetic sequences that start off functionless can gain a function eventually.

retrovirus is a parasite. so its origin need to be from the host genome and not the opossite

That doesn’t even make logical sense. Mosquitoes are parasites - they don’t come from humans. The tongue eating louse is a parasite - it doesn’t come from fish.

we have evidence that the virus itself “infected” by the host

This doesn’t make sense either. Viruses can’t be infected by hosts.

we find erv(s?) that contradict the phylogenetic tree

Probably because they come from different branches on that tree.

What you haven’t addressed is the direct evidence that ERVs once started out as RNA parasites and have left clear signs of integration wherever they have inserted themselves. Specifically:

Specifically they almost all have identical LTRs which are formed in the reverse transcription process from RNA → DNA and they almost all have identical TSRs (target site repeats) - this is when a bit of the target site is duplicated when the ERV integrates itself into that position.

There are plenty of other ways that we know that these weren’t originally there but you will have enough trouble explaining these two features on their own so I will leave you with these for the time being.

so we dont have any evidence that molecular clock can be credible.

I have just demonstrated to you that the predictions we get from molecular clocks agree with the predictions we get from population genetics through ILS.

19

Thanks for your explanation, Ace. Since I do not have time to read twenty papers (or however many) right now, I will simply take your word on the calculation of splits, and the correction of 2my between divergences. Humans/chimps diverged from gorillas 10 my, and human from chimp about 8 my ago, we will assume for now.

Sarfati notes Haldane’s dilemma, which indicates that even for the very long time of 10 my, there is not enough time for the required bp changes to be fixed in the population.

Take a population of 100,000. Suppose a Darwinian scenario claims a male and female each receive a new beneficial mutation, which substitutes into the population in one generation. In other words, the mutation is claimed to go from two copies to 100,000 copies in one generation. This would require a reproduction rate of 50,000. In a sexual species, this would require that females average at least 100,000 births each. For an ape-human-like species, this is obviously not plausible. If evolution happened in this highly unrealistic manner continously for 10 million years, only 500,000 mutations could be substituted. (with a generation time of 20 years). …Under Haldane’s analysis, the maximum speed sustainable over long-periods would be “one substitution per 300 generations”. (Sarfati, Greatest Hoax on Earth, p.55)

This realistic rate would equal about 1700 beneficial mutations (selected and fixed) in 10 my. This is quite a difference from the necessary 30-120 mbp difference required in 8 my. Even the theoretical very optimistic and unrealistic rate of 500,000 bp in ten my does not give the required 30 mbp. For small beneficial mutations, the selection effect is also very small, so delays or prevents the fixing of the mutation. Only mutations with a large beneficial effect are likely to be fixed (selected for) more quickly.

Some of this DNA will be functional and some will be non-functional but this is irrelevant to the matter at hand.

You say this, but earlier, Dennis asked the question, “Why would a designer independently create three species with this pattern when there is no need, biologically, for them to have this pattern?” But if it is functional dna, then obviously there would be a need to have the pattern, so it is not entirely irrelevant.

You mention ERVs. “If the same virus of the same class and subclass with the same mutations inserted itself into an identical location in many different species leaving the same tell-tale markers of insertion in all of them then the only reasonable explanation is that this happened once in a common ancestor…” But your sentence is illogical. You say that if the virus inserted into many species, then it is only reasonable that this only happened once in a common ancestor. This is self-contradictory. First, if a virus could do this once, then it could do it twice, or a million times. Second, if it could do it in many different species, then those species did not have to be one common ancestor, but rather, a group of ancestors, and so this event cannot be delineated by time so easily. It could as easily have happened long after a divergence, since why would we not expect the same marker to be left behind in various species.

20

ace. i talk about a parasite that depend on is host. so its logical that the virus origin is from the host itself.

you said:

“viruses can’t be infected by hosts”-

they actually can acquired genes from the host. the v-src gene is a good example.

“probably because they come from different branches on that tree”-

they are actually claim that this is the result of ils.

you said:

“what you haven’t addressed is the direct evidence that ERVs once started out as RNA parasites and have left clear signs of integration wherever they have inserted themselves”-

again- the opossite can be the true- the virus “infacted” by the host.

you said that:

“I have just demonstrated to you that the predictions we get from molecular clocks agree with the predictions we get from population genetics through ILS”-

what prediction? give me a specific example.

21

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.