Human-Chimp Similarity: only 70%?

[gecoverr]

I have one question about this paper "Initial sequence of the chimpanzee genome and comparison with the human genome" concerning the level of human-chimp similarity. As we read in the Nucleotide Divergence section:
"Best reciprocal nucleotide-level alignments of the chimpanzee and human genomes cover ~2.4 gigabases (Gb) of high-quality sequence, including 89 Mb from chromosome X and 7.5 Mb from chromosome Y."
My question is: Why did they align only about 76% of human genome with chimp genome? Was it (as some creationists e.g. Jeffrey Tomkins imply) because of huge differences between these two species in the remaining 24% of the genome? Or is there some other reason?
For example here - http://www.evolutionnews.org/2011/09/critically_analyzing_the_argum051321.html - Richard Buggs claims that because only 76% of human genome can be aligned with the chimp's, the similarity between them falls down to 70%.
Or here - http://www.icr.org/article/human-chimp-similarities-common-ancestry/ - Jeffrey Tomkins claims that: "One of the main problems with a comparative evolutionary analysis between human and chimp DNA is that some of the most critical DNA sequence is often omitted from the scope of the analysis. Another problem is that only similar DNA sequences are selected for analysis. As a result, estimates of similarity become biased towards the high side. An inflated level of overall DNA sequence similarity between humans and chimps is then reported to the general public, which obviously supports the case for human evolution. Since most people are not equipped to investigate the details of DNA analysis, the data remains unchallenged."
I would be very grateful if someone could clarify this issue for me.

[hersheyh]

On Friday, August 31, 2012 9:33:37 AM UTC-4, gecoverr wrote:
> I have one question about this paper "Initial sequence of the chimpanzee genome and comparison with the human genome" concerning the level of human-chimp similarity.

You should always actually put a complete citation in, so that people trying to respond do not have to search to find the article. This one is freely available at:
http://www.nature.com/nature/journal/v437/n7055/full/nature04072.html

But you may have failed to understand the meaning of the first words of the article: "Initial sequence..."
Initial does not mean complete. More complete analyses are available.

As we read in the Nucleotide Divergence section:
>
> "Best reciprocal nucleotide-level alignments of the chimpanzee and human genomes cover ~2.4 gigabases (Gb) of high-quality sequence, including 89 Mb from chromosome X and 7.5 Mb from chromosome Y."
>
> My question is: Why did they align only about 76% of human genome with chimp genome? Was it (as some creationists e.g. Jeffrey Tomkins imply) because of huge differences between these two species in the remaining 24% of the genome? Or is there some other reason?

It was NOT because of huge differences between these two species in the remaining 24% of the genome. As they point out, there are, in addition to 'point' mutation differences, differences of up to thousands of nts due to deletions and insertions (see Fig. 6) for a total of about 90 Mb (million basepairs) difference. These are the result, however, of no more than a few hundred mutation events. In addition there are known inversions and translocations that have occurred (which changes the *arrangement* of nts but not necessarily the sequence. Again, these are examples of a very few mutational events causing a big difference if you were to examine million base pair chunks across the boundary. You would get a human sequence, for example, that would have homology with DNA in two different chimpanzee chromosomes.

But what is important for evolution is the number of mutational events, not whether the mutational event causes a big "sequence change" of little consequence or causes a big "phenotypic" difference due to a single base change.

> For example here - http://www.evolutionnews.org/2011/09/critically_analyzing_the_argum051321.html - Richard Buggs claims that because only 76% of human genome can be aligned with the chimp's, the similarity between them falls down to 70%.

Even *before* we could sequence DNA, we *knew* this was false. If you denature DNA from two species and allow it to reanneal and then do thermal melting temperature analysis, you get a measure of DNA similarity that is similar to what you see in Fig. 3 where you examine 1 Mb segments. You get a mean difference of about 1.2%.

>
> Or here - http://www.icr.org/article/human-chimp-similarities-common-ancestry/ - Jeffrey Tomkins claims that: "One of the main problems with a comparative evolutionary analysis between human and chimp DNA is that some of the most critical DNA sequence is often omitted from the scope of the analysis. Another problem is that only similar DNA sequences are selected for analysis. As a result, estimates of similarity become biased towards the high side. An inflated level of overall DNA sequence similarity between humans and chimps is then reported to the general public, which obviously supports the case for human evolution. Since most people are not equipped to investigate the details of DNA analysis, the data remains unchallenged."
>
> I would be very grateful if someone could clarify this issue for me.

If you count the 90 Mb difference as being due to 250-500 insertion/deletion events with the number of nts affected being biased by a few large mutational events and ignore translocations and inversions (at most a few dozen), you get around 5% total sequence difference rather than 1.2%. But the number of mutational events that produced all these changes (LL would only modestly increase the number of mutational event differences from the 30 million point mutation differences. All the large deletions/insertions and rearrangements amount to fewer than a few hundred *mutational* differences. The vast majority of indels are small (<700 bp). The total estimate of all such indel differences (large or small) is about 5 million "mutational events". This is biased strongly toward small events with a blip due to transposable units at around 300 bp. Humans have significantly more Alu transposable elements interspersed in their genome (7000 rather than 2300), but the sequences of these elements is still identifiable

as a transposable unit.

That said, the *only* way for someone to have claimed in 2005 that there is only 70% sequence identity between humans and chimps would be for them to ignore all previous data on DNA homology or to lie about what constitutes homology (for example, all the Alu sequences that transpose show homology to each other, but are present in different locations). It also helps to claim that we should consider a single indel of a 40,000 base pairs as a 40,000 times more important than 40,000 changes at single nt sites.

But the fact remains that nearly all of the 30-35 million mutational differences between humans and chimps are selectively neutral and due to drift. The phenotypically important mutational differences -- those that natural selection has more rapidly changed -- can *sometimes* be identified by evidence of more rapid change and *more* difference. Other selectively important mutations, those that only required a small change at the DNA level and, often, only a small change in timing or amount (one or two nt or mutational differences) may be hidden among these 30-35 million mutational differences.

Bottom line: In terms of mutational differences, the human and chimp genome differ at about 1.2-1.5% sites. About 14% of these site differences may be due to large to small indels.

[Kermit]

On 31 Aug, 06:33, gecoverr <gecov...@gmail.com> wrote:
> I have one question about this paper "Initial sequence of the chimpanzee genome and comparison with the human genome" concerning the level of human-chimp similarity. As we read in the Nucleotide Divergence section:
> "Best reciprocal nucleotide-level alignments of the chimpanzee and human genomes cover ~2.4 gigabases (Gb) of high-quality sequence, including 89 Mb from chromosome X and 7.5 Mb from chromosome Y."
> My question is: Why did they align only about 76% of human genome with chimp genome? Was it (as some creationists e.g. Jeffrey Tomkins imply) because of huge differences between these two species in the remaining 24% of the genome? Or is there some other reason?

> For example here -http://www.evolutionnews.org/2011/09/critically_analyzing_the_argum05...- Richard Buggs claims that because only 76% of human genome can be aligned with the chimp's, the similarity between them falls down to 70%.
> Or here -http://www.icr.org/article/human-chimp-similarities-common-ancestry/- Jeffrey Tomkins claims that: "One of the main problems with a comparative evolutionary analysis between human and chimp DNA is that some of the most critical DNA sequence is often omitted from the scope of the analysis. Another problem is that only similar DNA sequences are selected for analysis. As a result, estimates of similarity become biased towards the high side. An inflated level of overall DNA sequence similarity between humans and chimps is then reported to the general public, which obviously supports the case for human evolution. Since most people are not equipped to investigate the details of DNA analysis, the data remains unchallenged."

> I would be very grateful if someone could clarify this issue for me.

hersheyh explains this better than I could, but I want to add one
thing:

There are several ways used to measure DNA similarity. Whatever method
is used, if it is used for all, then life demonstrates a nested
hierarchy of genomes that matches the nested hierarchy of morphology.
However you measure it, we are more closely related to the chimps than
we are to the gorilla, and more to the gorilla than to the wolf, etc.

Kermit

[jonsbulk...@gmail.com]

The latest articles (2014) by Tomkins and others continue to show the large variance between chimps and humans. Rambunctious evolutionists overstepped the bounds of actual sequence data.

[John Harshman]

It's just confusion, probably intentional on the part of the
creationists, over multiple measures of similarity. The usual figure,
98.7% similarity, refers to the average similarity of all alignable,
orthologous sequences. This 70% figure is presumably counting the
proportion of orthologous sequences. You should understand that the bulk
of these non-orthologous sequences consist of highly repetitive Alu
elements, most of them inserted into the human genome since our
separation from chimps. Why Alu elements should be considered critical
is beyond me.

The more important point is that by whatever measure you use, as long as
it's the same measure in each case, we are closer to chimps than to any
other species. Now why should that be? And what about the differences
would be difficult to explain by ordinary evolutionary processes?

Where are these "latest articles (2014) by Tomkins"?

[alias Ernest Major]

There was enough context to guess where they were.

http://www.icr.org/article/7892/
http://www.icr.org/article/7856/

--
alias Ernest Major

[Walter Bushell]

In article <7tadnR5hZormhW3P...@giganews.com>,

John Harshman <jhar...@pacbell.net> wrote:

> The more important point is that by whatever measure you use, as long as
> it's the same measure in each case, we are closer to chimps than to any
> other species. Now why should that be? And what about the differences
> would be difficult to explain by ordinary evolutionary processes?

BTW: are we closer to chimps or bonobos? Or is it too close to call?

--
Never attribute to stupidity that which can be explained by greed. Me.

[erik simpson]

Chimps and bonobos diverged from each other more recently than humans diverged
from the chimp-human common ancestor, so we are equally related to both, at
least in evolutionary terms.

[broger...@gmail.com]

On Monday, February 3, 2014 4:56:33 PM UTC-5, jonsbulk...@gmail.com wrote:

>
> >
>
> > I would be very grateful if someone could clarify this issue for me.
>
>
>
> The latest articles (2014) by Tomkins and others continue to show the large variance between chimps and humans. Rambunctious evolutionists overstepped the bounds of actual sequence data.

What degree of similarity between chimps and humans is predicted by special creation? 0%, 5%, 25%, 70% 95%, 98%? How do you decide?

[John Harshman]

On 2/3/14 3:37 PM, Walter Bushell wrote:
> In article<7tadnR5hZormhW3P...@giganews.com>,
> John Harshman<jhar...@pacbell.net> wrote:
>
>> The more important point is that by whatever measure you use, as long as
>> it's the same measure in each case, we are closer to chimps than to any
>> other species. Now why should that be? And what about the differences
>> would be difficult to explain by ordinary evolutionary processes?
>
> BTW: are we closer to chimps or bonobos? Or is it too close to call?
>

Nobody has done a complete bonobo genome yet, to my knowledge. I bet
you'd need a huge sample of the genome to see any consistent difference
in distance. Phylogenetically, of course, we are equally close to both
of them.

[jillery]

AIUI bonobos are a relatively recent variation of chimpanzee. So
wouldn't that mean they are more closely related to each other than to
humans?

[John Harshman]

Yes. That's what he said. But that doesn't necessarily tell you about
genetic distance.

[John Harshman]

Horrendous. I particularly like the end of the second article: "Research
using three large datasets produced by the ENCODE project is now
underway at ICR for the purpose of addressing these questions. In a
concurrent study, I am also comparing human protein-coding regions to
those in chimpanzees. In combination, these new analyses will provide a
much more detailed picture of what makes humans unique and will further
demonstrate we are not evolved apes."

In other words, we're in the early stages of research, and here are the
conclusions we're going to come to after we do it. Now that's science!

[Metspitzer]

On Mon, 03 Feb 2014 14:30:19 -0800, John Harshman
<jhar...@pacbell.net> wrote:

>On 2/3/14 1:56 PM, jonsbulk...@gmail.com wrote:
> > On Friday, August 31, 2012 6:33:37 AM UTC-7, gecoverr wrote:
> >> I have one question about this paper "Initial sequence of the
>chimpanzee genome and comparison with the human genome" concerning the
>level of human-chimp similarity. As we read in the Nucleotide Divergence
>section:
> >>
> >> "Best reciprocal nucleotide-level alignments of the chimpanzee and

>human genomes cover ~2.4 ?gigabases (Gb) of high-quality sequence,
>including 89? Mb from chromosome X and 7.5?Mb from chromosome Y."

> >>
> >> My question is: Why did they align only about 76% of human genome
>with chimp genome? Was it (as some creationists e.g. Jeffrey Tomkins
>imply) because of huge differences between these two species in the
>remaining 24% of the genome? Or is there some other reason?
> >>
> >> For example here -
>http://www.evolutionnews.org/2011/09/critically_analyzing_the_argum051321.html
>- Richard Buggs claims that because only 76% of human genome can be
>aligned with the chimp's, the similarity between them falls down to 70%.
> >>
> >> Or here -
> >> http://www.icr.org/article/human-chimp-similarities-common-ancestry/
> >> - Jeffrey Tomkins claims that: "One of the main problems with a
> >> comparative evolutionary analysis between human and chimp DNA is
> >> that some of the most critical DNA sequence is often omitted from

Why would some of the most critical DNA sequence be omitted?

[John Harshman]

Presumably, so as to arrive at the preselected result of 98% similarity.
Of course that's not true.

[Mr. B1ack]

On Mon, 3 Feb 2014 13:56:33 -0800 (PST), jonsbulk...@gmail.com
wrote:

>On Friday, August 31, 2012 6:33:37 AM UTC-7, gecoverr wrote:
>> I have one question about this paper "Initial sequence of the chimpanzee genome and comparison with the human genome" concerning the level of human-chimp similarity. As we read in the Nucleotide Divergence section:
>>

>> "Best reciprocal nucleotide-level alignments of the chimpanzee and human genomes cover ~2.4?gigabases (Gb) of high-quality sequence, including 89?Mb from chromosome X and 7.5?Mb from chromosome Y."

>>
>> My question is: Why did they align only about 76% of human genome with chimp genome? Was it (as some creationists e.g. Jeffrey Tomkins imply) because of huge differences between these two species in the remaining 24% of the genome? Or is there some other reason?
>>
>> For example here - http://www.evolutionnews.org/2011/09/critically_analyzing_the_argum051321.html - Richard Buggs claims that because only 76% of human genome can be aligned with the chimp's, the similarity between them falls down to 70%.
>>
>> Or here - http://www.icr.org/article/human-chimp-similarities-common-ancestry/ - Jeffrey Tomkins claims that: "One of the main problems with a comparative evolutionary analysis between human and chimp DNA is that some of the most critical DNA sequence is often omitted from the scope of the analysis. Another problem is that only similar DNA sequences are selected for analysis. As a result, estimates of similarity become biased towards the high side. An inflated level of overall DNA sequence similarity between humans and chimps is then reported to the general public, which obviously supports the case for human evolution. Since most people are not equipped to investigate the details of DNA analysis, the data remains unchallenged."
>>
>> I would be very grateful if someone could clarify this issue for me.
>
>The latest articles (2014) by Tomkins and others continue to show the large variance between chimps and humans. Rambunctious evolutionists overstepped the bounds of actual sequence data.

Um ... wouldn't matter if there was only one percent
similarity, it would demonstrate that humans are no
"special creation", something entirely distinct, but
just the 99 millionth re-arangement of the same old
stuff.

As for any quantitative figure of "alignment" ... I'd be
interested in seeing how "identical" a random sample
of humans actually are. There's really no such thing
as a "standard human" after all.

[Roger Shrubber]

Actually, it looks exactly like the type of thing congress would
be most comfortable funding: research being too risky unless you
can guarantee what you're going to be finding.

[jillery]

No, he said we are equally related to both, which is not what I said.

[Robert Carnegie]

Is there then any doubt that this is a case of creationists
making up whatever story they want to tell, with no fear of
scrutiny or criticism? Except from here, of course.

[John Harshman]

If you draw a tree, you will see that the two are equivalent.

[John Harshman]

On 2/3/14 7:12 PM, Mr. B1ack wrote:

> As for any quantitative figure of "alignment" ... I'd be
> interested in seeing how "identical" a random sample
> of humans actually are. There's really no such thing
> as a "standard human" after all.

It depends on how you count. The standard method is to look at SNPs and
to count the mean percent difference between two randomly chosen humans,
which comes to around 99.9% identical. If you consider indels, and count
each unmatched base as a difference, you would get a considerably larger
number, perhaps even as little as 98% identical, though I don't have any
hard data.

[jillery]

On Mon, 03 Feb 2014 23:22:27 -0800, John Harshman

Ok, I have it now. Modern humans are equally related to both modern
chimpanzee and modern bonobo. Thank you for pointing that out.

[Paul J Gans]

Walter Bushell <pr...@panix.com> wrote:
>In article <7tadnR5hZormhW3P...@giganews.com>,
> John Harshman <jhar...@pacbell.net> wrote:

>> The more important point is that by whatever measure you use, as long as
>> it's the same measure in each case, we are closer to chimps than to any
>> other species. Now why should that be? And what about the differences
>> would be difficult to explain by ordinary evolutionary processes?

>BTW: are we closer to chimps or bonobos? Or is it too close to call?

Depends whose DNA you look at. Congresscritters are closer to
chimps. Pop idols are closer to bonobos.

--
--- Paul J. Gans

[Paul J Gans]

Yes, but we don't speak to that branch of the family. Their table
manners are awful.

[Paul J Gans]

John Harshman <jhar...@pacbell.net> wrote:
>On 2/3/14 2:47 PM, alias Ernest Major wrote:
> > On 03/02/2014 22:30, John Harshman wrote:
> >> On 2/3/14 1:56 PM, jonsbulk...@gmail.com wrote:
> >> > On Friday, August 31, 2012 6:33:37 AM UTC-7, gecoverr wrote:
> >> >> I have one question about this paper "Initial sequence of the
> >> chimpanzee genome and comparison with the human genome" concerning the
> >> level of human-chimp similarity. As we read in the Nucleotide Divergence
> >> section:
> >> >>
> >> >> "Best reciprocal nucleotide-level alignments of the chimpanzee and

> >> human genomes cover ~2.4 ?gigabases (Gb) of high-quality sequence,
> >> including 89? Mb from chromosome X and 7.5?Mb from chromosome Y."

It makes paper-writing much easier.

[jillery]

Wasn't there someone named Sonny Bonobo?

[jillery]

But they're great at singles parties.

[RonO]

It is worse than these articles. It is really unbelievable what he
actually did to get his numbers.

Answers Research Journal 4 (2011):233–241.
www.answersingenesis.org/contents/379/arj/v4/human_chimpanzee_genome.pdf

This guy took chimp trace files (pretty much raw sequence data from an
automated sequence reads with minimal editing like removal of vector
sequence) and compared it to the human genome builds. He then claims
that his similarity measures mean something. His average similarity was
around 87% (range around 86-89% for sequences that aligned). Anyone
that has worked with trace sequences know that they are loaded with
errors. First there is the dye blob that you have to deal with where
around a hundred bases into the sequence you hit the region where
contaminant dyes migrate and mess up your sequence for around half a
dozen base reads. Next the vector editor is wrong quite often and
leaves vector sequence in if there are enough sequencing errors early or
late in the sequence (the most machine errors occur at the beginning of
the sequence and the end of the sequence where the sequencing vector
will be found). So you have to figure out what to trim. The best way
is to try to make alignments of the traces and see where things mess up.
Since he didn't do that he should have blasted the chimp trace
sequences against the chimp genome and looked at the match. The
expectation is that when he does this the match will only be a couple
percent better and he will know what parts of the sequence to ignore
(the dye blob) and what parts to trim (the error prone parts). When the
sequence is only 88 to 90% similar to the chimp that was sequenced that
should tell the boob just how wrong he is.

This paper is so pathetic that you have to wonder how long it took him
to figure out how to do something so wrong.

Ron Okimoto

[John Harshman]

What's just as weird is that he could have looked at the actual
human-chimp alignment at the UCSC genome database to directly refute his
claim that data were "cherry-picked". This is science with fingers in
the ears, shouting "la la la, I can't hear you".

[jillery]

How does one use something like the USC genome database to makes
comparisons between different species?

[John Harshman]

Rather than try to run my own tutorial, let me just direct you to the
site itself, which is well documented. But one thing the site contains
is a full alignment between the published human and chimp genomes.

http://genome.ucsc.edu/

[Paul J Gans]

Yup. Right on!

[jillery]

On Tue, 04 Feb 2014 13:20:34 -0800, John Harshman

I've already been there, and did not find what you say is there, which
is why I asked my question. I didn't expect a tutorial, but did hope
for something more specific than the home page.

[John Harshman]

OK. I don't do this very often, and they've redone the interface a
little since I last did. But you can first go to the chimp genome, and
then click on the human chain track, and then view parts in browser
window, and finally go to the bottom of that page ("together), and you
will see the aligned sequences. Granted, this doesn't show you the
entire alignment in one piece, but you can certainly view the entire
alignment a chunk at a time.

[John Harshman]

Or if you want the whole thing in a single chunk, this:

http://hgdownload.soe.ucsc.edu/downloads.html