Human evolution News

Evolutionary convergence saves creationist hypothesis over GULO

Spread the love

When similarities cannot be explained by common descent, evolutionists will plead “convergence” (some mysterious coincidental occurrence to explain similarity). Well what is good for the goose is good for the gander. Creationists are now realizing if evolutionists can claim convergence, so can creationists!

Evidence of Evolutionary Discontinuity and Genetic Entropy

AbstractModern genomics provides the ability to screen the DNA of a wide variety of organisms to scrutinize broken metabolic pathways. This wealth of data has revealed wide-spread genetic entropy in human and other genomes. Loss of the vitamin C pathway due to deletions in the GULO (L-gulonolactone oxidase) gene has been detected in humans, apes, guinea pigs, bats, mice, rats, pigs, and passerine birds. Contrary to the popularized claims of some evolutionists and neo-creationists, patterns of GULO degradation are taxonomically restricted and fail to support macroevolution. Current research and data reported here show that multiple GULO exon losses in human, chimpanzee, and gorilla occurred independently in each taxon and are associated with regions containing a wide variety of transposable element fragments. Thus, they are another example of sequence deletions occurring via unequal recombination associated with transposable element repeats. The 28,800 base human GULO region is only 84% and 87% identical compared to chimpanzee and gorilla, respectively. The 13,000 bases preceding the human GULO gene, which corresponds to the putative area of loss for at least two major exons, is only 68% and 73% identical to chimpanzee and gorilla, respectively. These DNA similarities are inconsistent with predictions of the common ancestry paradigm. Further, gorilla is considerably more similar to human in this region than chimpanzee—negating the inferred order of phylogeny. Taxonomically restricted gene degradation events are emerging as a common theme associated with genetic entropy and systematic discontinuity, not macroevolution.

HT
JoeCoder, tjguy

10 Replies to “Evolutionary convergence saves creationist hypothesis over GULO

  1. 1
    Jehu says:

    That is a great paper by Tomkins. As I recall Behe uses GULO to support common descent in Edge of Evolution. Didn’t Tomkin also do some research regarding the supposed chromosome fusion region in human DNA?

  2. 2
    RWalk says:

    Tomkins paper about “Alleged Human Chromosome 2 ‘Fusion Site’…”
    http://www.answersingenesis.or.....ome-fusion

  3. 3
    Aceofspades25 says:

    I was astounded to read this “paper” by Jeffrey Tomkins. At first I was astounded because I couldn’t believe that GULO had lost 6 exons independently in Humans, Chimpanzees, Gorillas and Orangutan. Then after checking up on the data I was astounded that this author would make so many obvious errors (either intentionally or unintentionally) (especially when the data is available to the public for verification).

    Right in the abstract the author makes the astounding claim that:

    The 28,800 base human GULO region is only 84% and 87% identical compared to chimpanzee and gorilla, respectively

    So the first thing I did was fetch the 28,800 base region that he was talking about from UCSC. I then blasted this sequence against two other human genomes, a chimpanzee, a bonobo, a gorilla and an orangutan using the NCBI blast tool. As expected his numbers were far off from the real figures. The blast search for chimpanzees found the sequence and reported that it was 97.5% identical (it takes into account gaps due to indels). After downloading that complete region of chimpanzee chromosome 8 and aligning it, out of 28067 complete positions, there are 522 SNPs making them 98.1% identical (this is effectively ignoring indels). See here for the chimpanzee sequences that the blast search matched and their associated similarity scores.

    The blast search for gorillas found the sequence and reported that it was 96.6% identical (it takes into account gaps due to indels). After downloading that complete region of gorilla chromosome 8 and aligning it, out of 28583 complete positions, there are 522 SNPs making them 98.2% identical (this is effectively ignoring indels). See here for the gorilla sequences that the blast search matched and their associated similarity scores.

    The blast search tells me that bonobos are 98% identical in this region. See here for the bonobo sequences that the blast search matched and their associated similarity scores.

    The blast search tells me that orangutans are about 94% identical in this region. See here for the orangutan sequences that the blast search matched and their associated similarity scores.

    Because I was able to compare three humans to the other apes, interestingly I was able to find locations where two of the humans had a mutation that the third didn’t, leaving the third human more similar to the other apes in this position (see exon 3). There is another position (15284) where two humans have a 3bp deletion that the third doesn’t, leaving the third more similar to Bonobos and Gorillas in this location. The humans were 99.9% identical to each other (counting only SNPs)

    A computer algorithm produced the following phylogenetic tree based on just this 28,800bp sequence alone.

    Like with most other genes and pseudogenes that produce similar phylogenetic trees this is exactly what we should expect. Once again, this is strong evidence for common ancestry. His claims are the exact opposite of what the data shows (especially when considering the full 28,800bp region)

    Here is the aligned sequence for 3 humans, a chimpanzee, a bonobo, a gorilla and an orangutan for the region the author was referring to. No need to take my claims at face value, browse this and verify them for yourself. Here is a link to a tool I wrote for doing this.

    Other things to note:
    4 base pairs deleted in the common ancestor to chimps, bonobos and the three humans at position 10781
    4 base pairs deleted in the common ancestor to chimps, bonobos and the three humans at position 10716
    4 base pairs deleted in the common ancestor to chimps, bonobos and the three humans at position 8499
    A point mutation in the common ancestor to chimps, bonobos and the three humans at position 3676

    There are many more features like this which is what leads the algorithm to deduce the expected phylogenetic tree.

    Things the author got wrong:

    Humans are more similar to chimps than gorillas in this 28,800bp region
    Humans and Chimps are 97.5% identical in this region (not 84%)
    Humans and gorillas are 96.6% identical in this region (not 87%)

    I am now going to try and understand how the author got results that were so incredibly misleading. He claimed he used a “previously established technique of optimized sequence slices and the BLASTN algorithm (Tomkins 2013b)”

    Here is the reference he gives for his method of comparing optimized sequence slices. It it seems to me that all he is doing is chopping up the sequence from humans into smaller chunks and then running a standard BLAST search on each of these chunks. This is effectively the same as what I have done except I didn’t chop the human sequence up into chunks first. If he had done this as he claims then that search would have clearly shown him that each of these chunks is between 97 and 98% identical. Dishonestly (in my opinion) he makes no mention of this inconvenient fact within his paper.

    So how did he get 84% similarity between humans and chimps in the 28,800bp region?

    Running a blast search on that region matches 97% of the query (There is a good reason for this: There is another small gap of unsequenced bases within the chimp genome here which explains the missing 3% – These unsequenced bases within the chimp genome are represented by the letter N. Naturally if something is unsequenced then it is not going to be found). The blast search also reports that these matched regions are 97% identical.

    I’m guessing he considered the fact that 97% of the region was matched (without looking into why – otherwise he would have discovered that 3% of the chimp sequence was unknown) together with the fact that 97% of this region was identical, multiplied these two fractions together and got a result of 94%. At this point, he either fudged his result and intentionally changed 94% to 84% or he misread his calculator and read off 84 instead of 94 (this seems unlikely to me because as you will see he makes exactly the same mistake with gorillas)

    So how did he get 87% similarity between humans and gorillas in the 28,800bp region?

    The blast search for gorillas covers 99% of the series (gorillas also have a tiny stretch of unsequenced positions here) and says they are 97% identical. 97% of 99% is 96%. I’m guessing that once again he either fudged this figure and changed it to 87% or once again he happened to misread his result.

    Either way, the simple reason he ended up with a larger number for gorillas than chimpanzees is because chimps have more unsequenced positions in this fragment than gorillas do.

    Also it is poor methodology to simply multiply out the %cover by the %identity. If there was a stretch of DNA that couldn’t be matched against, this could easily be due to a single insertion or deletion within either species. If there are 300 bases in humans that can’t be found in chimps, these could either have been inserted in humans in a single mutational event or deleted from chimps in a single mutational event. This is why we shouldn’t at all be considering the figure for %cover. I suspect he knew this but it was convenient for him to do this anyway because it artificially lowers the similarity between the two species.

    The other thing that should have rung alarm bells for him (and probably did) is this:

    He mentioned looking at the frequency of SNPs across the human GULO pseudogene. If he had done this, he would have found that 1.9% of the bases are polymorphic between humans and chimps (522 variable positions of the 28,067 positions that could be aligned) – the other type of difference that can arise are due to indels. In a given sequence, SNPs will vastly outnumber the number of indels, making the SNP rate a good first estimate of the similarity between two sequences. The actual number of mutations will never be more than double the number of SNPs – at most there should be 1 indel for every 4 or 5 SNPs.
    His mutation rate of 84% would imply a count of about 7500 mutational events. This is almost 15x the number of SNPs!

    So to summarise the things he got wrong in this part of the paper alone:

    1. He didn’t look into why 3% of the chimp sequence and 1% of the gorilla sequence couldn’t be matched against. If he had, he would have found that they are unsequenced.

    2. I’m guessing he multiplied the %cover by the %identity. This is poor methodology for determining the similarity between two sequences because a single mutational event could be responsible for a large change in the %cover.

    3. He either fudged his data or made a basic arithmetic mistake (0.97 * 0.97 = 0.94, not 0.84!).

    4. Even though he calculated the number of SNPs to be significantly lower than his number of mutations (nearly 15x lower!), this suspiciously didn’t set alarm bells ringing for him.

    Now onto his next mistake – He makes the claim that:

    The 13,000 bases preceding the human GULO gene, which corresponds to the putative area of loss for at least two major exons, is only 68% and 73% identical to chimpanzee and gorilla, respectively. These DNA similarities are inconsistent with predictions of the common ancestry paradigm. Further, gorilla is considerably more similar to human in this region than chimpanzee—negating the inferred order of phylogeny

    Since the author was far off with his first claim, it is worth checking up on this one as well.

    Once again a simple blast search reveals that within these 13,000 bases chimpanzees are 98% identical to humans. Here is a zip filegiving this result in html format. Gorillas are also 98% identical. Here is a zip file giving the result for gorilla in html format. You will notice that for both of these matches, a large central chunk hasn’t been matched – this is because this portion of the gorilla and chimpanzee genome is unknown. Here is the full 13,000bp sequence in chimps (notice the “N”s in the centre of the sequence). Here is the full 13,000bp sequence in gorillas (notice the “N”s in the centre of the sequence).

    Once again I assume he used his method of comparing optimized sequence slices. If he had done this then that search would have clearly shown him that each of these chunks is between 97 and 98% identical. Dishonestly (in my opinion) he makes no mention of this inconvenient fact within his paper and instead focuses on the fact that a BLAST search couldn’t find a match for 33% of the 13,000 bases within DZ1.

    He doesn’t look into why it couldn’t find a match. The answer of course is because large chunks in this region haven’t been sequenced for chimpanzees and gorillas. Here is some information on what those Ns mean. Note that 100 Ns doesn’t mean that there are exactly 100 bases that couldn’t be sequenced – rather it means that they estimate there to be about 100 bases that can’t be sequenced.

    This explains how he came up with the 68% figure. Once again he was counting large blocks that hadn’t been sequenced as differences. He was also being dishonest in not mentioning the fact that the sequences that could be aligned were on average 98% identical.

    I’m going to skip now to the end where he gives the 6 phylogenetic trees based on just the 6 remaining exons in humans.

    The most obvious thing to say about these exons is that these sequences are incredibly short and each contain only one or two SNPs. In each of these cases the degree of confidence would be nowhere near enough to deduce phylogenetic relationships.

    The next thing to say about these diagrams is that it is incredibly misleading to show two species branching off simultaneously when they are identical or both have a single divergence from humans in different places over a very short sequence. He does this is diagrams 1, 2 and 3.

    Finally his method of simply looking at the percentage difference from humans in order to deduce phylogenetic trees is a terrible one and is not at all reliable (especially when the sequences are this short. What he should be doing is looking for groupings that diverge from the ancestral sequence. For example if the ancestral sequence is C at position 10 and chimps and humans group together with a G at position 10, then that is a point in favour of chimps and humans having a common ancestor in which that mutation from C->G occurred.

    Note that I got the locations of these exons from the UCSC database – his chosen bounds for the 6 exons make his regions slightly larger than mine.

    The first diagram shows gorillas more similar to us than chimps. Here is the sequence that he uses to construct this phylogenetic tree. He bases this off a single mutation that occurs in a common ancestor to chimps and bonobos at position 53. The only thing that can really be deduced from this sequence (with very low confidence) is that Chimps and Bonobos are more closely related to each other than the other apes shown.

    The second diagram shows chimps and gorillas branching off together in red as if this tree contradicts the known relationship between these apes. Here is the sequence that he uses to construct this phylogenetic tree. He bases this off the fact that chimps and gorillas each have a single mutation but neglects to mention that this mutation happens in different positions and so couldn’t possibly imply relatedness. What this sequence does tell us is that the three humans all share a common ancestor (p15, reasonable confidence) and that gorillas and bonobos share a common ancestor that excludes chimpanzees (p43, low confidence).

    I’m sure most creationists believe that chimps and bonobos share a common ancestor so this illustrates quite nicely that occasionally groupings can be misleading. This either points to incomplete lineage sorting (most probable) or possibly that the chimp and gorilla each experienced this same point mutation independently.

    I was intrigued to look into the third exon since he claims it shows humans and gorillas are only 85% identical while humans and orangutan are 98.2% identical. Here is the sequence that he uses to construct this phylogenetic tree. The first glaring thing you will notice is that he counts a single 3bp deletion in gorillas as 3 independent mutations (just looking at this sequence will highlight how blatantly dishonest this is). This is entirely what leads him to draw his third bizarre phylogenetic tree. What can be deduced is that the three humans are more closely related to each other than the other apes (p15, reasonable confidence), and that two of the humans are more closely related to each other than the third (p45, low confidence). This is the shortest of the three exons and so is his most misleading dataset.

    The fourth diagram shows orangutan and gorillas out of order. Here is the sequence that he uses to construct this phylogenetic tree. I only count a single position where gorillas differ from humans and so I think he must have overstated the bounds of this exon and counted differences that don’t come into my sequence. What this shows us at position 49 is that bonobos, chimps and orang group together excluding gorillas and humans. Incomplete lineage sorting could explain this. What else may have happened here is that the ancestral sequence was C and both gorillas and humans happened to have mutation to a T or the ancestral sequence was a T and both orang and (chimps, bonobos) happened to have a mutation to a T. Position 72 shows a deletion that is common to all three humans and chimps.

    Here is the sequence that he uses to construct the 5th phylogenetic tree. It shows the 3 humans grouping together at one point and the chimp and bonobo grouping together at another point.

    Here is the sequence that he uses to construct the 6th phylogenetic tree. Contrary to what he claims, it doesn’t show orangutan more closely related to humans than gorillas. I don’t know what he means by exon 6 but his inferred relationship is nowhere close to what this sequence shows. At position 48 bonobos and chimps group together with low confidence.

    Contrary to what he claims, these sequences don’t contradict the known relationships between these species.

    I can’t see anywhere within this paper where the author justifies his remarkable claim that that GULO had lost 5 exons independently in Humans, Chimpanzees, Gorillas and Orangutan. More importantly the author fails to explain how creationism (as a scientific theory) can account for the same 5 exons (distributed randomly throughout the gene) independently going missing in all the apes studied including humans, chimpanzees, gorillas and orangutan.

    I have heard creationists argue that the reason why many remarkably similar mutations will happen in different species is because some parts of the sequence in question were prone to those.

    The problem is that this doesn’t explain why exactly the same exons have have gone missing in humans, neanderthals, denisovans, chimpanzees, bonobos, gorilla and orangutan because while this pseudogenisation occurred in the common ancestor to the haplorhini, other species have lost GULO independently and in those other species completely different exons have gone missing.

    Here is a diagram comparing the rat GULO gene to the GULO pseudogene in 5 other species. Notice how in the great apes, identical exons have gone missing (strongly suggesting that these went missing in a common ancestor), while in the guinea pig and the flying fox where GULO became a pseudogene independently, completely different exons have gone missing.

    I constructed this diagram out by obtaining sequences for the 12 rat exons and then blasting each of these against the sequence for humans, chimps, gorillas, guineas pigs and fruit bats. If you’re interested in the results I used to draw this diagram, they’re all here.

    Note: Exons shown in white with dotted borders are missing altogether. Exons shown in green have been matched with a high degree of confidence. Exons shown in orange are highly mutated and have only been matched with a low degree of confidence.

    – As expected, the three great apes are all missing the same exons: Exons 2, 3, 6, 8, 11

    – Guinea pigs are missing completely different exons: Exons 5 and 12

    – The fruit bat is missing exons 2, 9 and 10.

    This is exactly what we would expect if this gene became pseudogenised in the common ancestor to the three great apes and became pseudogenised independently in guinea pigs and flying foxes. This is exactly what we wouldn’t expect if we thought the reason that the great apes are missing the same exons is because tend to occur independently in the same spots.

    In conclusion, not only does the author provide misleading results, apply poor methodology and fail to justify some of his more remarkable claims but the evidence (when considering the entire length of the GULO sequence) points incredibly strongly towards shared ancestry and it confirms known phylogenetic trees. Appreciate once again that a computer algorithm produced this diagram by simply analysing the 28,800 base pairs that he bases this paper on.

    This “scientific paper” illustrates perfectly why we should stick to real scientific journals that use real peer review systems. Clearly the closest thing ARJ have to a peer review system is a “does this look good for creationism” system.

    I’m sorry to have to word this so strongly, but this paper is an embarrassment to both ARJ and the author and is an indictment of creation science.

    I would like to ask the admins of UD to allow me to publish this as an article on the site. Of course I would understand if they weren’t inclined to allow material to be published that challenges creationist perspectives.

  4. 4
    JoeCoder says:

    Although I disagree that a tree built from a single gene or pseudogene can be “strong evidence for common ancestry”, as an ID proponent who rejects common descent, I would also like to see Aceofspades25’s comment published as a top-level post.

    We’ve been discussing it on reddit.com/r/creation, but we’re a pretty small community and most of us don’t have much of a background in the biological sciences. I’m hoping if his work has wider exposure then more can comment on it’s validity–hopefully to draw the attention Tomkins himself for comment.

  5. 5
    Aceofspades25 says:

    Thanks JoeCoder

  6. 6
    RWalk says:

    @Ace

    “His mutation rate of 84% would imply a count of about 7500 mutational events. This is almost 15x the number of SNPs!”

    I assume by ‘mutation rate’ you mean the claimed similarity between the genomes. However, given that, your logic seems to be:

    1. assume the standard evolution story, thus the differences in the compared genomes must be due to mutation

    2. note Tomkin’s 84% similarity claim

    3. note evolutionary expectation: “In a given sequence, SNPs will vastly outnumber the number of indels”

    4. then point out that these 3 propositions lead to a contradiction: the number of SNP’s is much too low compared to the total required mutations

    However, while you conclude that the error is #2 above, it could just as logically be #1 and/or #3

  7. 7
    bornagain77 says:

    Rwalk, the next best thing to having Dr. Tomkins here to defend himself, is this very recent video interview he gave in which he revealed, like you, many of the highly questionable assumptions undergirding Darwinian thought in genome comparisons:

    The Myth of 98% Genetic Similarity between Humans and Chimps – Jeffrey Tomkins PhD. – May 2014 – video
    https://vimeo.com/95287522

  8. 8
    RWalk says:

    Thanks, @BA77

    I’ve seen that interview and other presentations made by Dr. Tomkins. As you noticed, I try to identify the driving assumptions made (and often unstated) within the arguments presented. It is most interesting to read the evolutionist arguments for what and why much of the genome is not included in the comparisons.

  9. 9
    jtomkins says:

    The BLASTN analyses done in this paper were performed after stripping all N’s from the data set and sequence slicing the large contiguous sequence into optimized slice sizes – all done on a local server using optimized algorithm parameters. My data not only takes into account gaps, but sequences present in human and absent in chimp, and vice versa. Doing an amateur armchair analysis on the BLAST web server with default parameters never designed for a one-on-one large scale genomic regional comparison as noted in the comment above by aceofspades25 is bogus. Of course, if the paper was actually read in it’s entirety in regards to the above comments this would have been obvious.

    Also, as noted in several evolutionary papers, which I cited in my paper, the large scale comparison and major differences in structural variability surrounding the GULO regions between humans and great apes in the intronic areas has been noted before. Interesting that the misleading post by aceofspades25 did not make note of that. My paper was in fact accurate in all respects and true to previous findings published by evolutionist themselves. My work just hashed out and exposed what was already known, but never previously elaborated upon because it shows just another aspect of what a complete fraud the human evolution paradigm truly is.

  10. 10
    bornagain77 says:

    Dr. Jeffrey Tomkins, thank you so much for taking the time to clear that up. ,,, Darwinists have been trying to diss your work ever since it came out. I have to admit that I did not know which data set to believe, theirs or yours, until I watched your interview with Ian Juby. I then knew that you are a man who knows exactly what he is doing in this field and who takes his research seriously. Thanks again.

Comments are closed.