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.
I welcome thoughtful comments and questions about this post, or about the topic of this series.
a challenge to this hypothesis of âshare mutationsâ will be the case of prestin gene:
the prestin gene share 14 mutations without a commondescent. so this case disprove any claim of share mutations=commondescent.
another problem is that we actually find contradict data for the phylogeny in this article:
"By contrast, humans share at least 28 unique physical characteristics with orangutans but only 2 with chimps and 7 with gorillas, the authors say. "
those 2 facts disprove the main claim in the article.
have a nice day
Thank you for an easily understood presentation. While in science we almost never refer to findings as proven or fact, the evidence is overwhelming that the process as presented or as presented with minor modifications is consistent with your conclusions.
Indeed the more we learn the more we see the glory of God in the workings of nature, be it on the molecular level or the astronomic.
Hi dcscccc,
Iâve actually written about the prestin protein before, which you might find informative:
http://biologos.org/blog/evolution-basics-convergent-evolution-and-deep-homology
The brief answer is that the prestin protein in bats and whales has been shaped, through natural selection, in two lineages to have similar (but not identical) features. Prestin is used in the auditory systems of all mammals, and in whales and bats this protein came under selection for use in echolocation in both lineages. We understand the mechanism, and itâs not at all surprising. Note that although the mutations are similar in the two lineages, they are not identical.
Your second objection is based on a paper that was widely panned back in 2009, and flew in the face of the evidence even then. The authors ignored genetic data, and instead tried to build a case on character traits like hair length (things we know can change easily in a short time frame). Itâs a wonder it was even published, but it goes to show that in science even somewhat odd ideas will be entertained, even if they donât fit the data of the time. What counts, however, is how the different models fare with new data - and the publication of the gorilla and orangutan genomes completely demolished that paperâs conclusions. At the DNA level, humans are most closely related to chimpanzees, then to gorillas, and then to orangutans.
Thank you, Dennis. This is one of the best blog series Iâve read on the evidence for common ancestry from genetics, and how evolution acts on populations, not individuals. Your comparisons to the evolution of language and your explanations of genetic mutations and nested hierarchies are building a solid framework for thinking about these things. Iâm looking forward to the rest of the series. Please continue to post articles in this style so the concepts are accessible to people without deep expertise. I appreciate it when scientists can translate for the rest of us in such an engaging way. Keep it up!
hi again dr venema.
actually the prestin gene do have identical mutations between some bats and dolphins:
âNevertheless, dolphins and porpoises share at least 14 derived amino acid sites in prestin with echolocating bats, including 10 shared with the highly specialised CF bats 2 and 3.â
so this evidence disprove any claim of share mutations=commondescent.
about the primate phylogeny- the problem is that even scientists dont know who is close to who. why we check this by genetic similarity and not by morphological similarity? according to this logic- even if fish and chimp was closest from genetic prespective, we will need to say in this case that chimp is more closer to fish then human?
Dr. Venema did not claim that âshare mutations=commondescentâ because that would be silly. His post is about patterns of shared mutations that inactivate genes. The paper you cite is about shared mutations that do not inactivate genes, and that perfectly correlate with physiology and behaviour. The paper provides evidence that the mutations were targets of positive selection. This case is a very, very rare example of presumed convergent evolution at the amino acid level. It is misleading to compare it to the olfactory receptor pseudogene story.
Convergent evolution is interesting but not surprising. Seeing it at the amino acid level is also interesting, but only surprising because it suggests that the adaptation (echolocation) depends on very specific protein shapes/configurations. In many other cases of convergence, the changes are different between the convergent lineages, suggesting that there are many ways to acquire the convergent phenotype. In the Prestin case, it looks like there are very few ways to get there.
You seem to think that the mutations are âthe same,â but the authors of the paper only claim that there are identical mutations at the amino acid level. Have you looked at the DNA sequences? The accession numbers are provided in the paper.
actually its not matter if its in the amino acid level or dna. its still a lots of share mutations. here is another examples of share mutations in the dna level:
so the main point is that share mutations\phylogeny of pseudogenes doesnt mean commondescent.
dcscccc,
It does matter whether you are talking about the DNA level or amino acids. There are often multiple possible changes in the DNA codon that can give the same amino acid. Itâs the DNA that matters for this argument.
The main point is that your position isnât determined by the evidence. Itâs determined by your pre-committment to special creation of species. If thatâs what you think you need to believe, fine, but you should admit to yourself that a literal interpretation of the Biblical account is the reason for your belief, not the evidence, because the evidence wonât lead to the conclusion that you want.
There arenât just a few shared mutations between species like humans, chimps and gorillas. There are millions of them, and thatâs counting only the complex mutations like sizable insertions and deletions. And its not just in pseudogenes. There are shared complex mutations all over the place, sometimes even in coding regions. Of course each species also has some species-specific mutations that have occurred since they diverged. For transposon insertions there are a few thousand that are specific to each species, but millions that are shared, which indicates a very long shared history before they diverged. And occasionally you will see simple point mutations that did occur twice independently, say in human and gorilla but not chimp, but they are a very low percentage of the total.
If you are really interested, itâs not that hard to use the genome browser web sites. Get the human sequence, pick a segment at random and look at the corresponding chimp and gorilla sequences in the same region. For simplicity, it should be a region that isnât part of a large segmental duplication. If you want to see a sample of that kind of thing, look at the figure at The Art of the Soluble: Transposable Elements and Common Descent of Humans and other Primates
pgarrison.
first- there is 20 amino acid and only 4 dna bases. so the ratio is1\5. its mean that 14 shared amino acid is like 70 shared dna bases!
2)my position determined by the evidence and not by belief. im actually former evolutionist. what is the evidence? i will give you 2 of them:
1)we know that a self replicat motor need a de signer. the flagellum is a self replicat motor. then the flagellum need a designer.
2)there is no step wise from complex system to another. so if we will have a self replicat car (for example)â it will not evolve into somthing like f15. so there is not step wise also in living systems that are no less complex.
you said:
"And occasionally you will see simple point mutations that did occur twice independently, say in human and gorilla but not chimp, but they are a very low percentage of the total. "
actually there is milions of them:
Hi dcscccc,
Youâve missed the point of Pgarrisonâs explanation of your mistake on the shared DNA vs shared amino acids. 14 amino acids in common does not mean 14 x 3 shared DNA bases - far from it. There are many ways to code for most of the amino acids - for example, the amino acid leucine can be coded for by six possible codons: CTT, CTA, CTG, CTC, TTA or TTG. In cases of convergent evolution (independent mutations leading to the same features) we see convergence at the amino acid level but often different DNA sequences underneath, as expected by independent events.
The post you linked to claiming repeated mutations in GULO is also wrong. The original paper (done by a group in Japan with little to no expertise in evolutionary biology) claimed that there were many repeated mutations. They were wrong - not because they did the sequencing wrong, but because they did the interpretation wrong. They assumed that the rat sequence was ancestral and that humans and guinea pigs independently mutated at several locations to the same DNA letter. In reality, the shared DNA sequences between human and guinea pig were ancestral, and the rat had mutated away from the ancestral sequence. Their erroneous conclusions were picked up and trumpeted by young and old-earth creationists (such as Reasons to Believe), but theyâre simply wrong. The original study was published in a Japanese journal of vitaminology - hardly a reputable source for information on evolution.
You can read more about the GULO shared mutation issue, and why the anti-evolutionary claims based on it are wrong here:
http://biologos.org/blog/a-tale-of-three-creationists-part-3
The other examples claimed in your sources (such as the remainder of the examples in the creation.com article, and the Discovery Institute / ID article, are also understood and expected when one understands that evolution is a population-level phenomenon. The claim is that some DNA sequences in humans do not most closely match chimpanzees, but rather other apes (e.g. gorillas). One example in the creation.com article has a sequence where gorillas and chimps match more closely than either does to human.
These examples are real, but the interpretation offered (that they are the result of independent mutations, or that they confound phylogenies) is incorrect. These are the expected results of what happens when large populations undergo speciation. Have you noticed that these examples invariably use another very close relative to humans (i.e. other apes)? There is a reason for this - the lineages leading to present-day humans, chimpanzees and gorillas separated from each other in a relatively short timespan, and did so as large populations (around 50,000 for each common population at the point of separation). As a result , we expect this sort of thing - if we didnât see it it would be a problem for evolution, actually.
The effect producing these patterns is called incomplete lineage sorting. We will actually be covering this next in this series in the context of using it to determine population sizes. Iâve also written about it before, which you might find helpful:
http://biologos.org/blog/evolution-basics-species-trees-gene-trees-and-incomplete-lineage-sorting
When you cover this again, it may be a good idea to include the maths (the theory behind this (theory of coalescent trees) was worked out in 1982) which predicts:
That for Humans, Chimps and Gorillas:
This is almost exactly what they found with the publication of the Gorilla genome
thanks for your response dr venema.
you said:
:â14amino acids in common does not mean 14 x 3 shared DNA bases - far from it. There are many ways to code for most of the amino acids -â-
yes, i know that. but the chance is the same. the chance for specific amino acid is 1\20. the chance for specific nucleotied is 1\4. so from pure chance prespective, without any selection involve, is more rare to get the same amino acid compare to the same dna base. simple calculationâŚ
âIn cases of convergent evolution (independent mutations leading to the same features) we see convergence at the amino acid level but often different DNA sequences underneath, as expected by independent eventsâ-
not always. see here for example:
. Genes tell story of birdsong and human speech -- ScienceDaily
ââThis means that vocal learning birds and humans are more similar to each other for these genes in song and speech brain areas than other birds and primates are to them,â Jarvis said.â
about the gulo- you said:
âThey assumed that the rat sequence was ancestral and that humans and guinea pigs independently mutated at several locations to the same DNA letter. In reality, the shared DNA sequences between human and guinea pig were ancestral, and the rat had mutated away from the ancestral sequence.â-
its unlikely for 3 reasons:
now about the ils model- first, if its true that for a 6 my split(human-chimp) we will get 15% different(like âaceofspadesâ claim)then its mean that for a split before about 50 my we will get a different of 100% in the genome. but we know that it is not true. even human and banana share near 50%. so somthing is wrong here.
secondly- with the ils model we can explain any phylogenetic contradiction. so how can we know when is convergent evolution and when it the ils?
about the evolution in general- again: we know that a gps cant evolve into a cell-phone(even if they will have a self replicat system). so we know that one kind of animal cant evolve into another.
yours sincerely
Hi dcscccc,
I donât know how to say this without sounding abrasive (Iâm not intending to be). You donât understand what you are reading, and taking quotes from a news article isnât going to be useful unless you understand the science behind it. The convergence between birds and humans with respect to speech/song is not at the DNA sequence level of individual genes. Itâs in how those genes, in general, are expressed (make into protein in certain times and places). There are similarities there, but the individual DNA sequences underlying those general similarities of expression are different. Many different DNA sequences can lead to (broadly similar) patterns of gene expression.
As for your comments on GULO, I canât really understand what you are arguing for. Note - I donât âassumeâ evolution. I accept the overwhelming evidence for common ancestry. Itâs not an assumption, itâs based on the evidence. Perhaps re-read the post I linked to on the topic.
Your question on ILS and phylogenies is a better one. We expect ILS for close relatives, and we expect a pattern (roughly) as follows:
The highest level of identity with our closest relative. (chimpanzees)
Based on population size and time since speciation, we predict a certain amount of our DNA will more closely match our next closest relative (gorilla).
This pattern will continue, with ever-smaller amounts of DNA matching more distantly-related organisms. For example, around 0.8% of our genome matches orangutans more closely than other great apes.
so, itâs not an arbitrary pattern, itâs a pattern that interlocks with other lines of evidence we have (we have the highest overall similarity to chimps, then gorillas, then orangutans, and so on).
Lastly, we do not have 50% genome identity with bananas. For some very, very highly-conserved genes, we might approach that figure at an amino acid level (not DNA), but not for an overall genome score.
Iâm not sure Iâll reply to other comments, but I hope you (and others) find this helpful.
hi again dr venema.
it may be my final response also
you said:
âThere are similarities there, but the individual DNA sequences underlying those general similarities of are different.â-
here is more clear example
or:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2690048/
âWe estimate that 113 convergent changes are distributed across all 13 mitochondrial protein-coding genes; this is â44 more convergent changes (presumably nonneutral) than predicted by empirically observed convergence levels between other branches over the entire tree, and â73 more than predicted by evolutionary model-based expectations.â
and again- even if it was in the amino acid level its not so different from dna level.
"As for your comments on GULO, I canât really understand what you are arguing for
i will try to explain it again. you claim that the rat sequance was changed and not the cavia sequance. so i explain that its very unlikely because the molecular âclockâ of the rat gulo is expact to be more slower then the cavia one. but there is no need for this because i show above a lots of convergent changes without commondescent.
. Note - I donât âassumeâ evolution. I accept the overwhelming evidence for common"
ancestry. Itâs not an assumption, itâs based on the evidence."-
science base on facts and not by belief. the fact is that we never seen an animal evolve a new complex systme like flagellum or eyes. according to the evolution its take milions of years. so its indeed a belief and not part of science because we cant test this claim in experiment.
âso, itâs not an arbitrary pattern, itâs a pattern that interlocks with other lines of evidence we have (we have the highest overall similarity to chimps, then gorillas, then orangutans, and so on).â-
this is again the phylogenetic tree. but like i said before- the morphological tree contradict the genetic one(and even the fossil one from what i remember). so according to what tree we should follow?
âLastly, we do not have 50% genome identity with bananasâ
not according to this source:
http://www.nhm.ac.uk/nature-online/evolution/what-is-the-evidence/morphology/dna-molecules/
"The genes of organisms that look very different are surprisingly similar. For example, human DNA sequences are over 95% identical to chimpanzee sequences and around 50% identical to banana sequences. "
do you have another source?
even so- according to the claim here of â15% predictionâ we will need to find a 100%changes in the genomes of 2 animals that split before 50 my⌠we know that isnt true- then its wrong.
have a nice day
dscccc, âthe molecular tree contradict the genetic oneâ
The molecular tree is the genetic tree. It is possible to infer trees for individual genes which sometimes donât agree with the tree inferred from whole genome sequences or from large sets of gene sequences for various reasons, including the incomplete lineage sorting mentioned by Dennis, but they are all molecular genetic trees.
You quote a paper which says in the title that the degree of convergence is unusual, so itâs not a general explanation for any similarity which is seen in different genomes. They made clear that the convergence seen was due to similar selection at high temperature, and that not all the mutations were convergent. That virus has a very small genome - only a few genes, which overlap with each other. As a consequence there are not that many possibilities for mutations to optimize for a new condition, so itâs not too surprising that some of them should repeat in an experiment with enough repetitions.
You donât think evolution happened at all, so why do you think convergence by selection somehow accounts for similarities in genomes? Even if it was all by selection, that wouldnât support your rejection of evolution. You would be invoking evolution to disprove evolution.
âso its indeed a belief and not part of science because we cant test this claim in experiment.â
Just before you made this statement you were arguing based on the observed sequences of mammals. Those were observations, not experiments. You need to make up your mind whether inferences from observations (based on understanding gained from experiments on present day organisms) are part of science or not. You canât have it both ways. Those of us who have actually been scientists are clear that this is part of science, and declaring that it isnât is just a rhetorical strategy, not a serious argument. You canât do experiments with stars, planets, black holes, etc. in astronomy - that doesnât keep it from being a science.
â1) the rat population is huge, so the fixation time for new mutation will take more time in the rat genome then the cavia oneâ
It doesnât matter. Huge numbers of neutral mutations happen in a population. A small proportion of them get fixed by chance. There has been plenty of time for fixation of a few neutral mutations in the GULO gene in rats.
ânot alwaysâ
âAlmost alwaysâ is what you need to account for the similarities between genomes. An occasional parallel occurrence of the same mutation in different species (the technical term is homoplasy) canât come anywhere near to accounting for what we see. I have seen a paper describing independent insertions of a transposon at the same site in 2 different species, but even then it wasnât the same kind of transposon, so the events were easily distinguished.
âeven so- according to the claim here of â15% predictionâ we will need to find a 100%changes in the genomes of 2 animals that split before 50 my⌠we know that isnt true- then its wrong.â
This isnât how this works. The multiplication isnât valid. The 15% is not a prediction based on how many mutations are expected in a given period of time. It has to do with the fact that two speciation events occurred close to each other in time, with the relative times estimated from coalescence theory.
Of course, mutations do accumulate with time. Mammals have a lot of similarities in sequences that have no functional effect, that are the result of recent common descent, but if you go as far as comparing a bird and a mammal there is little sequence similarity outside of coding and regulatory sequences. There are blocks of similar gene order, and of course itâs easy to detect the similarity in sequences that code for most RNAs and proteins and some regulatory sequences. Dennis was right that the mammalian and plant genomes are not 50% identical. That article was probably referring to the most highly conserved genes, but those are a very small portion of either genome.
Thereâs not any possible biological mechanism for the repeating of millions of complex mutations in multiple species. Targeted mutation canât do it, because enzymes are just not that accurate (and thereâs no evidence for it anyway.) Natural selection canât do it, because in complex organisms there are way too many ways for it accomplish similar functional results. Whichever one happens first is going to be selected.
I have to agree with Dennis that, if you are really interested in this stuff, you are going to have to go beyond news articles and creationist websites and take some college courses in molecular biology, or if you are capable of teaching yourself, get good textbooks in genetics, biochemistry, population genetics etc. and work your way through them.
I meant to add that I used to work with a family of proteins that was fairly well conserved and widely distributed in various organisms, including plants and microbes. If you looked at the gene from a mammal and compared it to that from plants or microbes, there would usually be ~20-25% identity at the protein level. 25% identity at the DNA level is what you what you expect for no relation at all (ignoring the effects of GC%.) When you make comparisons at those distances, the conservation is mostly due to functional requirements, since the common ancestor was over a billion years ago. That means a lot of mutations have happened at positions that donât have much functional constraint on them. In comparisons where the specific reaction catalyzed by the protein was different, the 3D structure of the two proteins would still be very similar, but you often couldnât detect any protein sequence similarity at all beyond a few positions that contributed directly to the catalysis.
âI have to agree with Dennis that, if you are really interested in this
stuff, you are going to have to go beyond news articles and creationist
websites and take some college courses in molecular biology, or if you
are capable of teaching yourself, get good textbooks in genetics,
biochemistry, population genetics etc. and work your way through them.â
Or, if youâre so inclined, you could start here:
hi pgarrison.
you said:
âThe molecular tree is the genetic tree.â-
yep. of course that i mean the morphologic one. i fix itâŚ
âYou quote a paper which says in the title that the degree of convergence is unusual, so itâs not a general explanation for any similarity which is seen in different genomes.â-
not realy. this article claim that is very unlikely that 12 mutaions will happen independenly in those sites. so i show that its very likelyâŚ
âJust before you made this statement you were arguing based on the observed sequences of mammals. Those were observations, not experiments.â-
its not part of evolution. one of the main part of the evolution theory is that complex system can evolve step wise. there is no experiment that show its possible. so all the experiments so far prove creation and not evolution.
ât. You canât do experiments with stars, planets, black holes, etc. in astronomy - that doesnât keep it from being a scienceâ-
yes. but we can test some of those field . the evolution theory we cantâŚ
" Huge numbers of neutral mutations happen in a population. A small proportion of them get fixed by chance. There has been plenty of time for fixation of a few neutral mutations in the GULO gene in rats"-
maybe. but we still have the same mutations in both cavia and human but not rat. so according to this- the rat clock is more speedy. but we know its not true because the rest of the exon is not share. so we have a problem here .
"Dennis was right that the mammalian and plant genomes are not 50% identical. That article was probably referring to the most highly conserved genes "-
maybe. do you have another source?.
âThereâs not any possible biological mechanism for the repeating of millions of complex mutations in multiple species.â-
first: who is talking about milions? secondly: if all the genomes made by the same designer- then even bilions bases can be in the same sites.
about books. i actually read a lots of themâŚ
âLet your conversation be always full of grace, seasoned with salt, so that you may know how to answer everyone.â -Colossians 4:6
This is a place for gracious dialogue about science and faith. Please read our FAQ/Guidelines before posting.