Chimeras and Horizontal Gene Transfer
Switch89
forum, and the subject came up of how Horizontal Gene Transfer affects
the Evolutionary Prediction that no true chimeras exist. The thread is
here:
http://www.freeratio.org/showthread.php?t=268272
Please cite papers or reliable resources if you intend to answer this
question.
John Harshman
the head of a goat, tail of a snake, and whatever else the Greek monster
had. So what do you mean? There are indeed organisms that are
combinations of characters from two species; they're called "hybrids".
If you mean a frog with the wings of a bat, even horizontal transfer
won't do that, and it requires a seriously naive view of genetics ("the
gene for bat wings") to imagine it would.
pol...@msx.dept-med.pitt.edu
A single gene transferred does not threaten anything - researchers
just use OTHER genes that don't move.
Horizontal transfers are readily detected and accounted for
(otherwise, how would anyone know it was an HGT ?)
This could be amusing though :
"Ancient origin of glycosyl hydrolase family 9 cellulase genes",
Davison and Blaxter
Molecular Biology and Evolution 22(5): 1273-84, 2005
Seems that many metazoans HAVE cellulase genes (many lineages have
lost them though), and that most examples of HGTs were WRONG !
Apparently the cellulase GHF9 found in Ciona is NOT an HGT; it was
just easier to find than those in other metazoans.
Klaus Hellnick
Chimeras DO exist. They are a result of grafting or fusion of zygotes.
rno...@umich.edu
wrote:
There are, indeed, human chimeras that carry some cells of one
parentage, other cells of a different parentage. These are quite
different from genetic hybrids where each cell carries some genes from
one parental stock, other genes from another. So if the subject is
horizontal gene transfer, the subject line is misleading as is the
question about chimeras.
Considering that viruses routinely inject their genetic material into
ours producing a commingled result, it is silly to insist that such
horizontal gene transfer cannot happen. Genetic engineering routinely
transfers specific genes from one organism into another. But, as John
Harshman pointed out, that does not mean that you get a fish with a
cow's head.
John Harshman
tunic made of cellulose? Are we sure that this gene acts as a cellulase
in most metazoans?
rno...@umich.edu
wrote:
There is some evidence for common origin of a variety of endogenous
animal cellulases suggesting that it was present in early bilateria.
See
Evidence for the presence of a cellulase gene in the last common
ancestor of bilaterian animals.
Nathan Lo, Hirofumi Watanabe, and Masahiro Sugimura
Proc Biol Sci. 2003 August 7; 270(Suppl 1): S69–S72
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1698037
John Harshman
> http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1698037
My question is whether these apparently homologous genes all act as
cellulases in the various taxa in which they occur.
Switch89
How are horizontal gene transfers readily detected? I mean, what is it
about a gene that reveals it to be the result of HGT?
John Harshman
course this only works if HGT isn't so common that there is no consensus
among other genes. But fortunately for us, HGT doesn't seem to be that
common. Even, I understand, in bacteria, but certainly so for eukaryotes.
Unholy Joe
Thank you,switch,for leading me here.
Gentlemen,
My points are thus:
1)Tunicates are urochordates,an urchin-chordate hybrid.Is it then too
far-fetched to postulate chimeras too?
2)John Harshman,you say that HGT can be identified by anomalous
phylogeny.But then,if you have a phylogeny of 80% noise,how can you
identify an anomaly?
3)In truth,how can you identify HGT at all till you get a phylogeny to
work with and considering there are two cases of anomalous rRNA aka
HGT,what makes you think that your "reference frame" phylogeny is
wrong?
4)In sum,how can you identify ancestry specifically at all?
P.S.I am not a creationist,nor an IDist.Rather I am an atheist.
P.P.S.I have sent all of the above to Francis Collins by email BTW.
Ernest Major
<0bb2031f-4ba0-448b...@c7g2000prc.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
>Ahhh,
>Thank you,switch,for leading me here.
>Gentlemen,
>My points are thus:
>1)Tunicates are urochordates,an urchin-chordate hybrid.Is it then too
>far-fetched to postulate chimeras too?
The concept that urochordates are urchin-chordate hybrids is a best an
extreme minority view. (In case you've been misled by the name, the
"uro" element in urochordate comes from the Greek "oura", meaning tail,
and refers to an anatomical element possibly homologous to the chordate
notochord.)
If one considers plant plastids, we find that the genes for elements of
the plastid proteome, whether encoded in the plastid or nuclear genome,
associate with bacteria in phylogenies, while genes for proteins active
in the cytosol associate with eukaryotes.
If urochordates were urchin-chordate hybrids we would expect to find
some genes to associate with urchins in phylogenies and some with
chordates. While I haven't looked at many deuterostome phylogenies I
would expect that such a striking result would be widely known.
Hybridisation is quite common in plants, and can be detected in
phylogenies in such fashion.
>2)John Harshman,you say that HGT can be identified by anomalous
>phylogeny.But then,if you have a phylogeny of 80% noise,how can you
>identify an anomaly?
That isn't a problem unless we actually have phylogenies of 80% noise.
>3)In truth,how can you identify HGT at all till you get a phylogeny to
>work with and considering there are two cases of anomalous rRNA aka
>HGT,what makes you think that your "reference frame" phylogeny is
>wrong?
>4)In sum,how can you identify ancestry specifically at all?
That's a non-sequitur. What you should be asking is how we can identify
relationships.
>P.S.I am not a creationist,nor an IDist.Rather I am an atheist.
>P.P.S.I have sent all of the above to Francis Collins by email BTW.
>
--
alias Ernest Major
Unholy Joe
> In message
> <0bb2031f-4ba0-448b-8daf-11a2b6543...@c7g2000prc.googlegroups.com>,
> Unholy Joe <transcendento...@gmail.com> writes
>
> >Ahhh,
> >Thank you,switch,for leading me here.
> >Gentlemen,
> >My points are thus:
> >1)Tunicates are urochordates,an urchin-chordate hybrid.Is it then too
> >far-fetched to postulate chimeras too?
>
> The concept that urochordates are urchin-chordate hybrids is a best an
> extreme minority view. (In case you've been misled by the name, the
> "uro" element in urochordate comes from the Greek "oura", meaning tail,
> and refers to an anatomical element possibly homologous to the chordate
> notochord.)
>
> If one considers plant plastids, we find that the genes for elements of
> the plastid proteome, whether encoded in the plastid or nuclear genome,
> associate with bacteria in phylogenies, while genes for proteins active
> in the cytosol associate with eukaryotes.
>
> If urochordates were urchin-chordate hybrids we would expect to find
> some genes to associate with urchins in phylogenies and some with
> chordates. While I haven't looked at many deuterostome phylogenies I
> would expect that such a striking result would be widely known.
>
> Hybridisation is quite common in plants, and can be detected in
> phylogenies in such fashion.
"In matters of science,the authority of a thousand is not worth the
humble reasoning of a single individual."
Urochordates have a genome 50-50 chordate and urchin-how do you
explain that?(from the new scientist article in case of sea squirts)
>
> >2)John Harshman,you say that HGT can be identified by anomalous
> >phylogeny.But then,if you have a phylogeny of 80% noise,how can you
> >identify an anomaly?
>
> That isn't a problem unless we actually have phylogenies of 80% noise.
prokaryote genes as due to HGT as well as 40 to 50 percent of human
genes as carried by viruses to the genomes (like ERVs)
>
> >3)In truth,how can you identify HGT at all till you get a phylogeny to
> >work with and considering there are two cases of anomalous rRNA aka
> >HGT,what makes you think that your "reference frame" phylogeny is
> >wrong?
> >4)In sum,how can you identify ancestry specifically at all?
>
> That's a non-sequitur. What you should be asking is how we can identify
> relationships.
rRNA itself can be subject to HGT,and how do you know that rRNA is not
anomalous(how can you determine its correctness with reference to a
tree,considering it is the first tree)?How can you determine whether
your reference frame is right or wrong?
Ernest Major
<4aa41f91-c9ef-4a37...@y33g2000prg.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
Firstly, how about a proper citation to the New Scientist article so we
can whether you've interpreted it correctly.
Secondly, how about a citation to the scientific literature, so we can
whether New Scientist has accurately presented the data.
>>
>> >2)John Harshman,you say that HGT can be identified by anomalous
>> >phylogeny.But then,if you have a phylogeny of 80% noise,how can you
>> >identify an anomaly?
>>
>> That isn't a problem unless we actually have phylogenies of 80% noise.
>I refer you again to the new scientist article which declares 80% of
>prokaryote genes as due to HGT as well as 40 to 50 percent of human
>genes as carried by viruses to the genomes (like ERVs)
You still haven't actually offered a reference to a New Scientist
article.
However, 80% of prokaryote genes being due to HGT is not the same as a
phylogeny of 80% noise.
>>
>> >3)In truth,how can you identify HGT at all till you get a phylogeny to
>> >work with and considering there are two cases of anomalous rRNA aka
>> >HGT,what makes you think that your "reference frame" phylogeny is
>> >wrong?
>> >4)In sum,how can you identify ancestry specifically at all?
>>
>> That's a non-sequitur. What you should be asking is how we can identify
>> relationships.
>Yeah well,that sums it up nicely.How can you get a relationship if
>rRNA itself can be subject to HGT,and how do you know that rRNA is not
>anomalous(how can you determine its correctness with reference to a
>tree,considering it is the first tree)?How can you determine whether
>your reference frame is right or wrong?
>>
>> >P.S.I am not a creationist,nor an IDist.Rather I am an atheist.
>> >P.P.S.I have sent all of the above to Francis Collins by email BTW.
>>
>> --
>> alias Ernest Major
>
--
alias Ernest Major
Unholy Joe
http://www.newscientist.com/article/mg20126921.600-why-darwin-was-wrong-about-the-tree-of-life.html
And 80% HGT=noise because of noncordant phylogeny.
Unholy Joe
Here's the new scientist article:
http://www.newscientist.com/article/mg20126921.600-why-darwin-was-wrong-about-the-tree-of-life.html
And yeah,80% HGT=noise because you'll get a noncordant phylogeny that
way.
Unholy Joe
Double post.
Unholy Joe
Here's the article:
http://www.newscientist.com/article/mg20126921.600-why-darwin-was-wrong-about-the-tree-of-life.html
And 80% HGT=noise.Noncordant phylogeny.
rno...@umich.edu
wrote:
> > http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1698037
>
> My question is whether these apparently homologous genes all act as
> cellulases in the various taxa in which they occur
Functional endogenous cellulase has been found in arthropods (insects
and crustacea), nematodes, and molluscs plus tunicates. That means
that the major branches of animal life -- deuterostome,
lophotrochozoa, and ecdysozoa -- all contain examples.
John Harshman
> Ahhh,
> Thank you,switch,for leading me here.
> Gentlemen,
> My points are thus:
> 1)Tunicates are urochordates,an urchin-chordate hybrid.Is it then too
> far-fetched to postulate chimeras too?
There is no urchin-chordate hybrid. That's a crackpot theory based on
unreplicable experiments.
> 2)John Harshman,you say that HGT can be identified by anomalous
> phylogeny.But then,if you have a phylogeny of 80% noise,how can you
> identify an anomaly?
By the 20% that isn't noise?
> 3)In truth,how can you identify HGT at all till you get a phylogeny to
> work with and considering there are two cases of anomalous rRNA aka
> HGT,what makes you think that your "reference frame" phylogeny is
> wrong?
You can't identify HGT till you get a phylogeny. I don't understand the
rest of the question.
> 4)In sum,how can you identify ancestry specifically at all?
By the usual methods of phylogenetic analysis. Sometimes it's easy, and
other times it's difficult. Consequently, sometimes identifying cases of
HGT is easy, and sometimes difficult.
> P.S.I am not a creationist,nor an IDist.Rather I am an atheist.
So?
> P.P.S.I have sent all of the above to Francis Collins by email BTW.
Also so?
John Harshman
>>> http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1698037
>> My question is whether these apparently homologous genes all act as
>> cellulases in the various taxa in which they occur
>
> Functional endogenous cellulase has been found in arthropods (insects
> and crustacea), nematodes, and molluscs plus tunicates. That means
> that the major branches of animal life -- deuterostome,
> lophotrochozoa, and ecdysozoa -- all contain examples.
>
rno...@umich.edu
> >>> http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1698037
> >> My question is whether these apparently homologous genes all act as
> >> cellulases in the various taxa in which they occur
>
> > Functional endogenous cellulase has been found in arthropods (insects
> > and crustacea), nematodes, and molluscs plus tunicates. That means
> > that the major branches of animal life -- deuterostome,
> > lophotrochozoa, and ecdysozoa -- all contain examples.
>
> While true, that doesn't answer the question
It answers it as far as the data go. There do exist functional
cellulase genes in a wide variety of animals. Perhaps a dozen species
have been studied. But the paper I cited was a gene database search
and obviously does not contain data about whether the genes are
functional or, if so, produce a functional cellulase.
Demonstrating that an animal contains a functional endogenous
cellulase is difficult as it is necessary to eliminate any possible
bacterial or protistan contamination. That such exist is an important
discovery. It is not really worthwhile to find every example because
we already know that bacterial and protistan cellulases to be
widespread and present in a tremendous range of guts. So it really
isn't that big a deal just what the percentage of cellulose you digest
is from your own enzyme as opposed to those of your internal friends.
Robert Carnegie
> Firstly, how about a proper citation to the New Scientist article so we
> can whether you've interpreted it correctly.
>
> Secondly, how about a citation to the scientific literature, so we can
> whether New Scientist has accurately presented the data.
New Scientist is bullshit even when it's correct. But what about
mitochondria? I don't know what the current status is of the idea
that they are descended from complete independent living cells
imported into some ancestor of ours, but if so, it's not just
horizontal gene transfer, it's horizontal entire-entity transfer. And
is outside many simple descriptions of evolution.
But that only means that some descriptions of evolution maybe should
be less simple than they are. And I say "maybe" because sometimes an
explanation that is oversimplified and doesn't cover the whole picture
is better than not understanding at all.
Unholy Joe
First off,the sea squirt has an 80% chordate genome.(a simple googling
of sea squirt genome)That makes new scientist's 50-50 figure WRONG.But
hold it...the sea squirt does have some urchin genes.Common descent or
hybridism?Oh...I just can't tell them anymore.
As for John's response of 80% noise,although what you say is
correct,and I would agree with you,a creationist or IDiot will see it
as common design rather than descent based on the overwhelming
noise.Whatcha say to that?
To point 3)My point is that since phylogeny has to be identified
before identification of HGT,you cannot detect an HGT that occurs in
your tree if you analyze a single gene like the 16s or 23s rRNA.If
rRNA itself has undergone HGT you cannot detect it for it is rRNA you
use for tree construction.Enter whole-genome analysis confounded by
80% HGT.BTW,if you look at the new web of life you will spot some
localized transfers as well as random distant transfers(look at the
left part of the web).If you have a pattern like that you might get a
concordant(because of localization) but wrong phylogeny because of
HGT.
To point 4)And do we not know that HGT is a potential confounding
factor in that?
P.S.I think talkorigins should have a page on this(the website,not the
google groups)
P.P.S.The T.O. creationist claims index has one faulty rebuttal(well
ok the contradictory evidence is very recent,it's about 3.4 billion
years old iron).
Unholy Joe
have it too.But I am unsure of how that will affect the genomic
analysis of sea squirts and their relationship with chordates.
Ernest Major
<1ab80c79-6063-497a...@z5g2000vba.googlegroups.com>,
Robert Carnegie <rja.ca...@excite.com> writes
>On May 21, 10:46 am, Ernest Major <{$t...@meden.demon.co.uk> wrote:
>> Firstly, how about a proper citation to the New Scientist article so we
>> can whether you've interpreted it correctly.
>>
>> Secondly, how about a citation to the scientific literature, so we can
>> whether New Scientist has accurately presented the data.
>
>New Scientist is bullshit even when it's correct. But what about
>mitochondria? I don't know what the current status is of the idea
>that they are descended from complete independent living cells
>imported into some ancestor of ours, but if so, it's not just
>horizontal gene transfer, it's horizontal entire-entity transfer. And
>is outside many simple descriptions of evolution.
I was about to say that we all know that horizontal transfer occurs, but
one second thoughts let me rephrase it as the existence of horizontal
transfer is well known among biologists and interested laymen.
However, the request was for a citation for tunicates (or more precisely
Ciona intestinalis) having a genome "50-50 chordate and urchin". I would
have preferred you to have retained that context.
>
>But that only means that some descriptions of evolution maybe should
>be less simple than they are. And I say "maybe" because sometimes an
>explanation that is oversimplified and doesn't cover the whole picture
>is better than not understanding at all.
>
--
alias Ernest Major
rno...@umich.edu
It will affect the relationship of sea squirts exactly zero.
You should know that humans and chicken share about 50% of their
genes. We also share an awful lot of genes (I can't find the exact
number just now) with plants. That sea urchins and tunicates might
share a percentage of genes is totally unsurprising. They both run
pretty much the same biochemistry with the same cellular machinery.
John Harshman
John Harshman
> Harshman,
> First off,the sea squirt has an 80% chordate genome.(a simple googling
> of sea squirt genome)That makes new scientist's 50-50 figure WRONG.But
> hold it...the sea squirt does have some urchin genes.Common descent or
> hybridism?Oh...I just can't tell them anymore.
Where do you get the 80/20% claim?
> As for John's response of 80% noise,although what you say is
> correct,and I would agree with you,a creationist or IDiot will see it
> as common design rather than descent based on the overwhelming
> noise.Whatcha say to that?
I say that it hardly matters what a creationist will see, since he sees
whatever he wants to. Common design doesn't explain nested hierarchy,
except to the degree that it explains anything you could conceivably see
and thus explains nothing.
> To point 3)My point is that since phylogeny has to be identified
> before identification of HGT,you cannot detect an HGT that occurs in
> your tree if you analyze a single gene like the 16s or 23s rRNA.
That depends on whether you have a prior idea of the real tree, gathered
from other data.
> If
> rRNA itself has undergone HGT you cannot detect it for it is rRNA you
> use for tree construction.
Agreed, if that was indeed what you did.
> Enter whole-genome analysis confounded by
> 80% HGT.BTW,if you look at the new web of life you will spot some
> localized transfers as well as random distant transfers(look at the
> left part of the web).If you have a pattern like that you might get a
> concordant(because of localization) but wrong phylogeny because of
> HGT.
What left part of what web? Are you allergic to citations?
> To point 4)And do we not know that HGT is a potential confounding
> factor in that?
In what? Are you also allergic to quoting the post you're responding to?
While we're at it, why did you reply to me but attach the reply to a
post by Robert Carnegie?
> P.S.I think talkorigins should have a page on this(the website,not the
> google groups)
> P.P.S.The T.O. creationist claims index has one faulty rebuttal(well
> ok the contradictory evidence is very recent,it's about 3.4 billion
> years old iron).
Why not start a new thread about that? Be specific.
pol...@msx.dept-med.pitt.edu
> Harshman,
> First off,the sea squirt has an 80% chordate genome.(a simple googling
> of sea squirt genome)That makes new scientist's 50-50 figure WRONG.
Actually, the sea squirt has about 16,000 genes, 80% of which are
similar to those in humans.
Given that humans are also chordates, this is not overly surprising.
That 80% is the SIGNAL that would be used to generate phylogenies, not
'noise' as you think it is.
> But
> hold it...the sea squirt does have some urchin genes.Common descent or
> hybridism?Oh...I just can't tell them anymore.
Common descent most likely, given the fact that researchers that
actually know what they are doing can tell the difference between
hybridism and common descent.
> As for John's response of 80% noise,although what you say is
> correct,and I would agree with you,a creationist or IDiot will see it
> as common design rather than descent based on the overwhelming
> noise.Whatcha say to that?
That 80% is the SIGNAL, not noise.
> To point 3)My point is that since phylogeny has to be identified
> before identification of HGT,you cannot detect an HGT that occurs in
> your tree if you analyze a single gene like the 16s or 23s rRNA.
Where did you get the idea that ribosomal RNAs transfer ?
Sane researchers use more than one gene to generate phylogenies if at
all possible.
>If rRNA itself has undergone HGT you cannot detect it for it is rRNA you
> use for tree construction.
Cite where anyone CLAIMED rRNA horizontally transfers ? And HOW would
they know ?
>Enter whole-genome analysis confounded by
> 80% HGT.BTW,if you look at the new web of life you will spot some
> localized transfers as well as random distant transfers(look at the
> left part of the web).If you have a pattern like that you might get a
> concordant(because of localization) but wrong phylogeny because of
> HGT.
Good thing that researchers that know what they're doing can detect
and account for HGTs.
> To point 4)And do we not know that HGT is a potential confounding
> factor in that?
Given that HGTs are detectable and can be accounted for and worked
around, it is a potential confounding factor, but not an overly
harmful one.
An earlier post of yours :
> The fact that it is a minority view matters little for a la Galileo:
> "In matters of science,the authority of a thousand is not worth the
> humble reasoning of a single individual."
Ah yes, the the standard 'Galileo Gambit'. Good thing EVIDENCE trumps
authority and data-free ruminations.
> Urochordates have a genome 50-50 chordate and urchin-how do you
> explain that?(from the new scientist article in case of sea squirts)
I suspect you misinterpreted something.
> > >2)John Harshman,you say that HGT can be identified by anomalous
> > >phylogeny.But then,if you have a phylogeny of 80% noise,how can you
> > >identify an anomaly?
> > That isn't a problem unless we actually have phylogenies of 80% noise.
> I refer you again to the new scientist article which declares 80% of
> prokaryote genes as due to HGT as well as 40 to 50 percent of human
> genes as carried by viruses to the genomes (like ERVs)
Actually, the article stated that 80% of 161 prokaryotes showed signs
of having undergone a few HGTs, not that 80% of all prokaryote genes
are HGTs.
40-50% of the human genome is composed of NON-INFECTIOUS (immobile)
ERVs, fragments of ERVs, replicating transposons, fragments of
transposons, and other detritus from millions of years of evolution.
The genes that are NOT part of infective viruses tend to stay put, and
those are the ones used to generate phylogenies.
> > >3)In truth,how can you identify HGT at all till you get a phylogeny to
> > >work with and considering there are two cases of anomalous rRNA aka
> > >HGT,what makes you think that your "reference frame" phylogeny is
> > >wrong?
What, EXACTLY, are these two cases of anomalous rRNA ? I suspect you
have misunderstood something.
> > >4)In sum,how can you identify ancestry specifically at all?
> > That's a non-sequitur. What you should be asking is how we can identify
> > relationships.
> Yeah well,that sums it up nicely.How can you get a relationship if
> rRNA itself can be subject to HGT,
Do you have a cite WHERE rRNA underwent an HGT ?
> and how do you know that rRNA is not
> anomalous(how can you determine its correctness with reference to a
> tree,considering it is the first tree)?How can you determine whether
> your reference frame is right or wrong?
By looking at several different genes - they HAVE derived trees for
many other genes you know.
An rRNA HGT would stick out like a sore thumb on the middle of your
forehead - a bacterial rRNA in a eukaryotic lineage would be quite
noticeable and require an explanation.
TomS
<smeRl.7426$Lr6....@flpi143.ffdc.sbc.com>, John Harshman stated..."
>
>Unholy Joe wrote:
>> Harshman,
>> First off,the sea squirt has an 80% chordate genome.(a simple googling
>> of sea squirt genome)That makes new scientist's 50-50 figure WRONG.But
>> hold it...the sea squirt does have some urchin genes.Common descent or
>> hybridism?Oh...I just can't tell them anymore.
>
>Where do you get the 80/20% claim?
>
>> As for John's response of 80% noise,although what you say is
>> correct,and I would agree with you,a creationist or IDiot will see it
>> as common design rather than descent based on the overwhelming
>> noise.Whatcha say to that?
>
>I say that it hardly matters what a creationist will see, since he sees
>whatever he wants to. Common design doesn't explain nested hierarchy,
>except to the degree that it explains anything you could conceivably see
>and thus explains nothing.
Like so much of creationist "explanations", at best all it does is
give temporary relief. If the creationists would ever think a little
bit about "common design", they'd see that it suggests some probably
uncomfortable things about "design" and the "designers".
But the actual behavior of the creationists is not to think about
their "explanations", so "common design" is just a slogan to get out
of a difficult situation.
--
---Tom S.
"As scarce as truth is, the supply has always been in excess of the demand."
attributed to Josh Billings
Ernest Major
<jharshman....@pacbell.net> writes
>Unholy Joe wrote:
>> Harshman,
>> First off,the sea squirt has an 80% chordate genome.(a simple googling
>> of sea squirt genome)That makes new scientist's 50-50 figure WRONG.But
>> hold it...the sea squirt does have some urchin genes.Common descent or
>> hybridism?Oh...I just can't tell them anymore.
>
>Where do you get the 80/20% claim?
A simple googling of sea squirt genome finds several pages reporting
that 80% of Ciona genes are also found in vertebrates. That observation
would seem to be orthogonal to a Michael Syvanen's claim of a 50-50
division between Ciona genes closer to vertebrates versus genes closer
to sea urchins - a large number of Ciona genes will also be found in
echinoderms. The natural reading of the 80% is that the other 20% are
genes lost from vertebrates, genes novel in tunicates (including any
horizontally transferred genes that tunicates have picked up), and genes
that have diverged beyond recognition.
http://www.universityofcalifornia.edu/news/article/5047
My first suspicion would be that Michael Syvanen's results are a
combination of long branch artefacts, saturation artefacts, and sparse
taxon sampling. But Google Scholar doesn't turn up any papers relating
to that study, so it's rather difficult to evaluate it.
--
alias Ernest Major
Unholy Joe
another 12 hrs or so.Reason:I am busy.(But not running away,oh no!)
P.S.I must say I am encouraged by the vitriol concerning lack of
citations,etc.This vitriol is EXACTLY what leads to a really rigorous
SCIENCE discussion.You have a nice ecological niche for me,or so it
seems to me at least :-).Posting on non-scientifically rigorous forums
seems to have downgraded me.
P.P.S.That galileo quote was something of a runaway thought.I lifted
it out of a sort of scrapbook of famous scientists I have made over
the years.
You'll all get my next replies in about 12 hrs.One final note:The
anomalous rRNA cases are footnotes 55 and 56 of the paper Horizontal
gene transfer:a critical view.Also,if you have citation needs,please
google it first instead of asking,it is not conducive to a discussion
if you bawl over something that can be resolved by a simple search
engine.
pol...@msx.dept-med.pitt.edu
In other words, *I* must do *YOUR* work for you. Do you always assume
everyone has access to everything you do - you are the one making
claims, so the onus is on you to back it up (and not expect others to
do your research for you).
From "Horizontal Gene Transfer : A Critical Review" :
"Only two partial transfers of rRNA are known among the thousands of
examples of organisms for which rRNA sequences are available (55, 56).
This, along with the observation that transgenic bacteria containing
rRNA operons from very closely related organisms are debilitated (53),
suggests that rRNA is an unusually stable phylogenetic marker. Indeed,
there are no indications that transgenic exchanges of rRNA domains are
as common in nature as suggested by Gogarten et al. (10). All in all,
the available data suggest that rRNA-based phylogeny is robust and
that Darwinian lineages are the essence of phylogeny."
Reference 55 notes that a thermophilic archaebacterium has 2 different
rRNA genes - its native one plus a new one.
Reference 56 notes that a thermophilic actinomycete has a few copies
of rRNA from another thermophilic actinomycete.
No mention of one kind of rRNA replacing another. Ever. So your
fears of rRNA being transferred so rampantly that it throws into
question the validity of the tree seems unfounded.
Ernest Major
<e6cf50e4-a662-43f9...@x1g2000prh.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
intended references (sic) 55 and 56 in PNAS 100(17): 9658–9662 (2003),
which translate to
* Genetics 130, 399–410 (1992), and
* J. Bacteriol. 181, 5201–5209
Fortunately all three papers are open access.
The first reference reports the existence of two distinct 16S rRNAs in
halophilic archaebacterium Haloarcula marismortui. This is not obviously
a case of HGT - it looks more like a gene duplication to me, but it
could be HGT followed by (gene) chimaerism.
The second refererence reports the existence of two distinct types of
rRNA operons in the thermophilic actinomycete Thermomonospora
chromogena. They make a case for this being an instance of HGT, from
Thermobispora bispora. It also contains further references to papers
reporting duplicate rRNA genes in Dugesia mediterranea and Plasmodium
berghei.
* Mol. Biol. Evol. 13: 824–832 (1996)
* Science 238: 933–937 (1987)
* Int. Rev. Cytol. 99: 295–309 (1986)
The first is accessible. The phylogenetic tree suggests the possibility
of horizontal transfer, but the second sequence is not expressed in
Dugesia mediterranea, and may be a pseudogene, in which case the tree
could also be explained by the accelerated sequence change of
pseudogenes.
The second is paywalled, and I can't find the third. However elsewhere I
find a mention of 2 RNA genes in Plasmodium cynomolgi and 3 in
Plasmodium vivax. This looks like a case of gene duplication.
* J. Mol. Biol. 269, 203-213 (1997)
--
alias Ernest Major
rno...@umich.edu
<snip>
>....One final note:The
> anomalous rRNA cases are footnotes 55 and 56 of the paper Horizontal
> gene transfer:a critical view.Also,if you have citation needs,please
> google it first instead of asking,it is not conducive to a discussion
> if you bawl over something that can be resolved by a simple search
> engine.
It is a courtesy to provide enough information for us to find a
specific paper you refer to. At the minimum, you should provide first
author name, journal, and year. Better is a complete citation with
journal, year, volume, and pages. The title alone is much less
helpful. Professionals refer to papers by the author(s) name(s), not
by title.
It is simply rude to force your readers to try to search for a
specific paper by guessing at the proper key words to use in google
and then searching page after page for something that looks right.
But this is usenet and rudeness is commonplace, though always
remaining rude.
It is also to be noted that
Horizontal gene transfer: A critical view
C. G. Kurland, B. Canback, and Otto G. Berg
PNAS August 19, 2003 vol. 100 no. 17 9658-9662
http://www.pnas.org/content/100/17/9658.abstract
(which is how I prefer to cite papers)
says in the abstract, copied here in totality:
It has been suggested that horizontal gene transfer (HGT) is the
“essence of phylogeny.” In contrast, much data suggest that this is an
exaggeration resulting in part from a reliance on inadequate methods
to identify HGT events. In addition, the assumption that HGT is a
ubiquitous influence throughout evolution is questionable. Instead,
rampant global HGT is likely to have been relevant only to primitive
genomes. In modern organisms we suggest that both the range and
frequencies of HGT are constrained most often by selective barriers.
As a consequence those HGT events that do occur most often have little
influence on genome phylogeny. Although HGT does occur with important
evolutionary consequences, classical Darwinian lineages seem to be the
dominant mode of evolution for modern organisms.
In other words, just as we have been telling you, HGT is no big deal
for evolution except for primitive genomes.
rno...@umich.edu
> In message
<snip all context to leave just citation examples>
> PNAS 100(17): 9658–9662 (2003),
> * Genetics 130, 399–410 (1992), and
> * J. Bacteriol. 181, 5201–5209
> * Mol. Biol. Evol. 13: 824–832 (1996)
> * Science 238: 933–937 (1987)
> * Int. Rev. Cytol. 99: 295–309 (1986)
> * J. Mol. Biol. 269, 203-213 (1997)
I prefer some mention of at least the first author, as a courtesy to
the researcher(s), but these are clear, unambiguous citations that
anybody with the slightest training in science can very easily
recover. For brevity, this form can't be beat, except by using only
the first page instead of inclusive pagination.
Ernest Major
<5a67ec70-8c6a-4868...@e23g2000vbe.googlegroups.com>,
"rno...@umich.edu" <rno...@umich.edu> writes
sufficient for Google Scholar to find the papers.
--
alias Ernest Major
spintronic
wrote:
> Switch89 wrote:
> Its phylogeny doesn't match that of the other genes in the organism. Of
> course this only works if HGT isn't so common
Pfff.
> But fortunately for us, HGT doesn't seem to be that common.
Pfff.
> Even, I understand.
That is remarkable.
spintronic
> Ahhh,
> Thank you,switch,for leading me here.
> Gentlemen,
> My points are thus:
> 1)Tunicates are urochordates,an urchin-chordate hybrid.Is it then too
> far-fetched to postulate chimeras too?
No
> 2)John Harshman,you say that HGT can be identified by anomalous
> phylogeny.But then,if you have a phylogeny of 80% noise,how can you
> identify an anomaly?
They couldn't even identify contamination.
Apparently it's a personal choice of the researcher wether he passes
on the anomily
he discovers.
Most the time they just ignore it, and claim it's a glitch.
(Till it bite's them in the ass, and they can no longer ignore the
glitch)
> 3)In truth,how can you identify HGT at all till you get a phylogeny to
> work with and considering there are two cases of anomalous rRNA aka
> HGT,what makes you think that your "reference frame" phylogeny is
> wrong?
All depends where they discover the gene first.
Say they discover gene X in snakes. It's a snake gene.
Even if they find it in 10 other critters.
> 4)In sum,how can you identify ancestry specifically at all?
Easy, they make it up.
> P.S.I am not a creationist,
Neither am I.
> P.P.S.I have sent all of the above to Francis Collins by email BTW.
I have took my complaint to the IACXC.
spintronic
> They don't offer citations unfortunately.
> And yeah,80% HGT=noise because you'll get a noncordant phylogeny that
> way.
Citations are there, you Just have to get your "pipe, violin &
magnifying glass out".
rno...@umich.edu
> "rnor...@umich.edu" <rnor...@umich.edu> writes
>
>
>
> >On May 21, 3:11 pm, Ernest Major <{$t...@meden.demon.co.uk> wrote:
> >> In message
>
> ><snip all context to leave just citation examples>
>
> >> PNAS 100(17): 9658–9662 (2003),
> >> * Genetics 130, 399–410 (1992), and
> >> * J. Bacteriol. 181, 5201–5209
> >> * Mol. Biol. Evol. 13: 824–832 (1996)
> >> * Science 238: 933–937 (1987)
> >> * Int. Rev. Cytol. 99: 295–309 (1986)
> >> * J. Mol. Biol. 269, 203-213 (1997)
>
> >I prefer some mention of at least the first author, as a courtesy to
> >the researcher(s), but these are clear, unambiguous citations that
> >anybody with the slightest training in science can very easily
> >recover. For brevity, this form can't be beat, except by using only
> >the first page instead of inclusive pagination.
>
> I though I was being a little unkind - but it turns out that this is
> sufficient for Google Scholar to find the papers.
I do it the old fashioned way -- trudge through the library stacks to
find the hard-copy journal, then copy down the relevant details on 4x6
index cards (the 3x5 are too small to hold enough information).
No, that was years ago. Does anybody need a whole pile of very nice
metal index card boxes? Now I just go directly to the journal web
site to search by year/volume/page. Of course it helps that my
institutional affiliation gives me free web access to full text of
just about every journal!
John Harshman
[M]adman
How brilliant a creator to reuse so much of the same genes yet still achieve
such dramatic differences with each life form.
Switch89
be found, how does this prediction hold up in light of our knowledge
of horizontal gene transfer?
If we found something with half mammalian DNA and half avian DNA,
could that be chalked up to Horizontal Gene transfer? Why or why not?
Boikat
Who? Gay God? The God that was lonely and created a guy, and not a
woman, to be his companion?
Boikat
Unholy Joe
Madman,please go away.I don't like your antics.If you have something
valuable to say,then SAY it.
>Citations are there, you Just have to get your "pipe, violin &
>magnifying glass out".
What is the meaning of this?
> 1)Tunicates are urochordates,an urchin-chordate hybrid.Is it then too
> far-fetched to postulate chimeras too?
>No
The urchin-chordate story has fallen apart with a little digging.The
rest of spintronic's post seems to be slandering rather than honest
inquiry of science methods.
Right,onto HGT and rRNA:
Look,how could they identify it if somehow one organism replaced its
rRNA with HGT and the donor died?The tree itself is built upon rRNA so
how could they identify such a case?Wouldn't the new rRNA look like
the organism had it all along?And thus a wrong tree with a wrong
organism in a wrong taxon.
I hope I have clarified.
That's my principal question too!Answers?
My own spin on it is can transitional fossils be ascribed to HGT?
Can,for example,be tiktaalik a result of tetrapod genes transferred to
fish?I would imagine that gene would not be expressed,but can it ever
happen even if it is not often and can creationists use it to explain
transitionals satisfactorily under their religion?
On the topic of the human genome and viruses:
>40-50% of the human genome is composed of NON-INFECTIOUS (immobile)
>ERVs, fragments of ERVs, replicating transposons, fragments of
>transposons, and other detritus from millions of years of evolution.
>The genes that are NOT part of infective viruses tend to stay put, and
>those are the ones used to generate phylogenies.
And these portions came into the genome..when?Cambrian?Precambrian?
Whenever they did they would mess up the phylogeny.Say they came
during the Cambrian when the Chordate phylum was emerging.And they
came from,oh I don't know,Mollusca.So now you have Chordates grouping
up with molluscs due to the transfer when they wouldn't otherwise.
Your tree gets messed up.
>An rRNA HGT would stick out like a sore thumb on the middle of your
>forehead - a bacterial rRNA in a eukaryotic lineage would be quite
>noticeable and require an explanation.
Oh,but we aren't talking about that-just prokaryote to prokaryote
transfer maybe.And if you do judge correctness by accordance with the
tree by phenotype,then please note that phenotype is not always a
correct indicator of genetic similarity.And while we are at it,please
note that this accords with ID more than evolution-a designer has no
obligation to put stuff the right way.And while I am at it...
>By looking at several different genes - they HAVE derived trees for
>many other genes you know.
And are these not more prone to HGT?I understand rRNA is highly
conserved and would be a bit less prone to transfer.
>Actually, the article stated that 80% of 161 prokaryotes showed signs
>of having undergone a few HGTs, not that 80% of all prokaryote genes
>are HGTs.
Check out the original paper here:
http://www.pnas.org/content/105/29/10039
And read the abstract.
BTW,here's the web of life:
http://en.wikipedia.org/wiki/File:Horizental-gene-transfer.jpg
>Common design doesn't explain nested hierarchy,
>except to the degree that it explains anything you could conceivably see
>and thus explains nothing.
It is unfalsifiable,no doubt about that.But just note that a web is
not a nested hierarchy.If over 80% of genes is transferred you can't
be sure what is a taxon and what it falls under anymore.
Unholy Joe
That's a funny one.Perhaps god is bisexual.
Unholy Joe
Are you Calilasseia from the RD.forum?Cause your posting seems similar
to Cali's........
Ernest Major
<bbd3b107-ab02-4209...@g1g2000yqh.googlegroups.com>,
"rno...@umich.edu" <rno...@umich.edu> writes
don't know that for every journal, and I don't think that we can expect
the average poster to do so either. As it happens I picked up two of the
papers by going to the journal web sites, but it would have been quicker
using Google Scholar.
--
alias Ernest Major
Ernest Major
<29ab46c0-de87-477e...@d19g2000prh.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
>Replying...
Could you be so kind as to conform to local custom, and make your
responses to the post(s) you are replying to, and not to the latest
post.
>Madman,please go away.I don't like your antics.If you have something
>valuable to say,then SAY it.
>
>>Citations are there, you Just have to get your "pipe, violin &
>>magnifying glass out".
>
>What is the meaning of this?
It's a reference to Arthur Conan Doyle's character Sherlock Holmes. But
pay no attention to Spintronic - he's Madman's slightly cleverer
brother.
>
>> 1)Tunicates are urochordates,an urchin-chordate hybrid.Is it then too
>> far-fetched to postulate chimeras too?
>
>>No
>
>The urchin-chordate story has fallen apart with a little digging.The
>rest of spintronic's post seems to be slandering rather than honest
>inquiry of science methods.
>
>Right,onto HGT and rRNA:
>Look,how could they identify it if somehow one organism replaced its
>rRNA with HGT and the donor died?The tree itself is built upon rRNA so
>how could they identify such a case?Wouldn't the new rRNA look like
>the organism had it all along?And thus a wrong tree with a wrong
>organism in a wrong taxon.
>
>I hope I have clarified.
In the early days 18S RNA was a common gene used for phylogenetic
studies. It turns out that it's not an ideal locus for this purpose - I
am not sure of the details, but I think it's because its rate of
evolution varies drastically between lineages. Subsequently we have
looked at other loci.
We don't assume that the 18S RNA tree represents the true tree, and the
others loci are wrong. For example in the secondarily amitochondriate
microsporidia 18S makes them appear to be basal eukaryotes, while other
loci show them to be derived and simplified fungi. The former is a long
branch artefact. (Parasitic organisms often have accelerated rates of
sequence evolution.) Thus we could identify a case of HGT of 18S RNA by
its non-congruence with the majority of other loci, with the proviso
that it may not be possible to distinguish a sufficiently ancient HGT
event from a long branch artefact.
>
>>Here's the issue: If common descent predicts that no true chimeras may
>>be found, how does this prediction hold up in light of our knowledge
>>of horizontal gene transfer?
>
>>If we found something with half mammalian DNA and half avian DNA,
>>could that be chalked up to Horizontal Gene transfer? Why or why not?
>
>That's my principal question too!Answers?
Something with half-mammalian and half-avian DNA would be like a
pre-Cambrian rabbit - highly surprising. Until we actually find such a
thing speculation to how it could be explained is just speculation.
>
>My own spin on it is can transitional fossils be ascribed to HGT?
If transitional fossils are to be ascribed to HGT then one would expect
that living morphological intermediates would be genetic chimaeras.
Would you like to propose to the Discovery Institute that they fund that
sequencing of the genomes of, say, lungfish and coelocanths, lampreys
and hagfish, tree shrews, and so on, with a view to testing this
conjecture. Or perhaps you might like to consider the chronological
distribution of transitional fossils and the taxa you think that they're
the result of HGT between. The distribution in time of these taxa
doesn't support your spin.
>Can,for example,be tiktaalik a result of tetrapod genes transferred to
>fish?I would imagine that gene would not be expressed,but can it ever
>happen even if it is not often and can creationists use it to explain
>transitionals satisfactorily under their religion?
Some creationists have been observed to speculate, or perhaps even
claim, that assorted hominin fossils are the result of hybridisation
between men and apes.
One problem with the conjecture that Tiktaalik is the result of tetrapod
genes transferred to fish is that Tiktaalik is dated at 375 million
years old, and the earliest tetrapods at 365 million years old.
>
>On the topic of the human genome and viruses:
>>40-50% of the human genome is composed of NON-INFECTIOUS (immobile)
>>ERVs, fragments of ERVs, replicating transposons, fragments of
>>transposons, and other detritus from millions of years of evolution.
>>The genes that are NOT part of infective viruses tend to stay put, and
>>those are the ones used to generate phylogenies.
>
>And these portions came into the genome..when?Cambrian?Precambrian?
Try Oligocene or Miocene, if not more recently.
>Whenever they did they would mess up the phylogeny.Say they came
>during the Cambrian when the Chordate phylum was emerging.And they
>came from,oh I don't know,Mollusca.So now you have Chordates grouping
>up with molluscs due to the transfer when they wouldn't otherwise.
>Your tree gets messed up.
That wouldn't happen. Firstly because they're not that old, and secondly
because the great majority of that class of DNA is not conserved, and
therefore wouldn't give a tree at all when comparing metazoan phyla.
>
>>An rRNA HGT would stick out like a sore thumb on the middle of your
>>forehead - a bacterial rRNA in a eukaryotic lineage would be quite
>>noticeable and require an explanation.
>
>Oh,but we aren't talking about that-just prokaryote to prokaryote
>transfer maybe.
When you bring up the possibility of half-mammal half-avian DNA, and
claims that tunicates are chordate-urchin hybrids, we clearly aren't
talking about just prokaryote to prokaryote transfer. But a bacterial
rRNA in an archaeal lineage would be just as glaring an anomaly.
>And if you do judge correctness by accordance with the
>tree by phenotype,then please note that phenotype is not always a
>correct indicator of genetic similarity.And while we are at it,please
>note that this accords with ID more than evolution-a designer has no
>obligation to put stuff the right way.And while I am at it...
You said you were not a creationist. One wonders why a non-creationist
would make a comment about something according more with ID that with
evolution. To suggest that evolution results in a perfect correlation of
genotype and phenotype is a strawman - we were aware of the existence of
phenotypic convergence before we even knew that DNA was the genetic
material. So is judging correctness with accordance with phenotype - we
just correctness by congruence among total evidence.
>
>>By looking at several different genes - they HAVE derived trees for
>>many other genes you know.
>
>And are these not more prone to HGT?I understand rRNA is highly
>conserved and would be a bit less prone to transfer.
That may be the case. But this is another point where evidence would not
go amiss.
>
>>Actually, the article stated that 80% of 161 prokaryotes showed signs
>>of having undergone a few HGTs, not that 80% of all prokaryote genes
>>are HGTs.
>
>Check out the original paper here:
>http://www.pnas.org/content/105/29/10039
>And read the abstract.
What the paper say appears to be between those two positions. It says
that 80% of gene families in prokaryotes have undergone horizontal
transfer at least once - this is not the same as 80% of the genes of an
individual prokaryote being the result of horizontal transfer. For a
hypothetical example, consider the existence of 8 lineages, where 10% of
the genome has been transferred from lineage A to lineage B, and so on
up 10% being transferred from lineage H to lineage A. If there are no
secondary transfers 80% of genes have undergone horizontal transfer, but
only 10% of the genome of each lineage has.
Note also that the paper says "Although there can be little doubt that a
considerable component of prokaryote genome evolution over recent
evolutionary time scales is fundamentally treelike in nature."
>
>BTW,here's the web of life:
>http://en.wikipedia.org/wiki/File:Horizental-gene-transfer.jpg
>
>>Common design doesn't explain nested hierarchy,
>>except to the degree that it explains anything you could conceivably see
>>and thus explains nothing.
>
>It is unfalsifiable,no doubt about that.But just note that a web is
>not a nested hierarchy.If over 80% of genes is transferred you can't
>be sure what is a taxon and what it falls under anymore.
>
Horizontal gene transfer is not the only confounding effect in attempts
to establish deep phylogenies. It may be that we can't resolve the
relationships between, say, the major bacterial clades. But your claim
that we can't identify taxa is wrong.
Consider a hypothetical example. Assume that we're looking at a sample
of 100 taxa and 10 loci, and that each has been horizontally transferred
once. (A 100% transfer rate in your terms.) The result is that in each
case of transfer the clade that received of the gene has a different
sister group than in the other 9 trees. The rest of the tree topology is
congruent in all ten cases.
--
alias Ernest Major
Greg Guarino
Your sentence above has suffered a random mutation.
Greg Guarino
rno...@umich.edu
Enough, I'll confess. I find the journal web site by googling it with
the name of the journal (or the organization for things like
"Science", "Nature", or "Evolution".) And the reason I don't just put
the citation directly into Google Scholar is because it never occured
to me to do that. But of course from now on, that will be the method
of choice so thank you for the information.
rno...@umich.edu
> On Thu, 21 May 2009 17:53:01 -0700, John Harshman
>
> >spintronic wrote:
> >> On May 21, 11:31 am, Unholy Joe <transcendento...@gmail.com> wrote:
> >>> Ernest,
> >>> Here's the new scientist article:http://www.newscientist.com/article/mg20126921.600-why-darwin-was-wro...
> >>> They don't offer citations unfortunately.
> >>> And yeah,80% HGT=noise because you'll get a noncordant phylogeny that
> >>> way.
>
> >> Citations are there, you Just have to get your "pipe, violin &
> >> magnifying glass out".
>
> >Ceci n'est un pipe.
>
> Your sentence above has suffered a random mutation.
Give him a free pas -- it's not his native language.
Nevertheless, I am surprised about his sexual preference in pipes.
Unholy Joe
Nuff said.
But before I go let me say that:
Considering common design is not being an IDist.Hear the phrase
"devil's advocate" much?
John Harshman
"Une"? Not only isn't it my native language, I don't speak it at all.
Therefore any approximation to a genuine sentence is a wonder and a marvel.
Would you prefer "Methinks it is not like a pipe"?
John Harshman
John Harshman
explanation, one that I do not so far summon up. One think it might
require is a world whose rules of organismal development were quite
different from our own. I'm reminded of the Simpsons episode in which
Lisa discovers a fossilized angel, and asks Steven J. Gould about it.
Hey, how about a Cambrian rabbit with 50% bumblebee genes?
rno...@umich.edu
wrote:
That completely destroys the meaning of the original, which is best
simply "this is not a pipe". The "this" in question is, of course,
not a pipe at all but rather a painting which contains an image
closely resembling a pipe plus the sentence in question. And, to be
more specific, the "this" that is ordinarily referred to is not even
one of the several similar original paintings but rather a printed
reproduction of a painting that contains an image resembling a pipe
or, worse, the computer reconstruction of a digital file containing
pixel values derived from a reproduction of a painting that contains
an image resembling a pipe.
Whether you think it is like a pipe is completely irrelevant.
TomS
<QnxRl.18880$D32....@flpi146.ffdc.sbc.com>, John Harshman stated..."
It wouldn't fly.
--
---Tom S.
"As scarce as truth is, the supply has always been in excess of the demand."
attributed to Josh Billings
TomS
<blxRl.22831$as4....@nlpi069.nbdc.sbc.com>, John Harshman stated..."
And how brilliant to use poor design principles in designing.
Any ordinary designer could use the similar designs to achieve
similar goals.
Greg Guarino
I hadn't noticed the gender confusion. I can jury-rig enough French to
get by while traveling, but I view the whole concept of "gender" for
clearly genderless objects as a pagan pantheist anachronism. For any
word as useless to me as "pipe" it's a tossup. In a pinch, I try to
work in a plural. :)
It was the deletion of the "pas" that caught my eye, hence rnorman's
play on mots, um, words.
>Would you prefer "Methinks it is not like a pipe"?
I'll cede the faux (I hope) Usenet brawl to rnorman on this point.
Greg Guarino
Unholy Joe
> Unholy Joe <transcendento...@gmail.com> writes
>
> >Replying...
>
> Could you be so kind as to conform to local custom, and make your
> responses to the post(s) you are replying to, and not to the latest
> post.
>
> >Madman,please go away.I don't like your antics.If you have something
> >valuable to say,then SAY it.
>
> >>Citations are there, you Just have to get your "pipe, violin &
> >>magnifying glass out".
>
> >What is the meaning of this?
>
> It's a reference to Arthur Conan Doyle's character Sherlock Holmes. But
> pay no attention to Spintronic - he's Madman's slightly cleverer
> brother.
A "consensus of the genes" tree,eh?But-correct me if I am wrong-what
if each gene gives you a slightly different in the lower groups yet
overall pretty similar phylogeny?You'll get this due to HGT and will
get some anomalies or "fuzzy edges".Would that be decided by
consensus,too?Or will the genes that have been congruent in
determining the higher relationship be used in constructing the lower
branches too?
All of them?That's one heck of a transfer!And that recently?I thought
HGT was principally occuring in the Archean and has abated a bit
since?
>
> >Whenever they did they would mess up the phylogeny.Say they came
> >during the Cambrian when the Chordate phylum was emerging.And they
> >came from,oh I don't know,Mollusca.So now you have Chordates grouping
> >up with molluscs due to the transfer when they wouldn't otherwise.
> >Your tree gets messed up.
>
> That wouldn't happen. Firstly because they're not that old, and secondly
> because the great majority of that class of DNA is not conserved, and
> therefore wouldn't give a tree at all when comparing metazoan phyla.
So other phyla likely don't even possess them?
>
>
>
> >>An rRNA HGT would stick out like a sore thumb on the middle of your
> >>forehead - a bacterial rRNA in a eukaryotic lineage would be quite
> >>noticeable and require an explanation.
>
> >Oh,but we aren't talking about that-just prokaryote to prokaryote
> >transfer maybe.
>
> When you bring up the possibility of half-mammal half-avian DNA, and
> claims that tunicates are chordate-urchin hybrids, we clearly aren't
> talking about just prokaryote to prokaryote transfer. But a bacterial
> rRNA in an archaeal lineage would be just as glaring an anomaly.
And that would be due to HGT,I suppose?
>
> >And if you do judge correctness by accordance with the
> >tree by phenotype,then please note that phenotype is not always a
> >correct indicator of genetic similarity.And while we are at it,please
> >note that this accords with ID more than evolution-a designer has no
> >obligation to put stuff the right way.And while I am at it...
>
> You said you were not a creationist. One wonders why a non-creationist
> would make a comment about something according more with ID that with
> evolution. To suggest that evolution results in a perfect correlation of
> genotype and phenotype is a strawman - we were aware of the existence of
> phenotypic convergence before we even knew that DNA was the genetic
> material. So is judging correctness with accordance with phenotype - we
> just correctness by congruence among total evidence.
Playing devil's advocate is an essential way to test theories.On
congruence with the total evidence,I can imagine it's a bit like
this,your suggestion of how phylogenies are made.
Gene 1 is an anomaly on one node.All other genes speak otherwise.So
gene 1 loses node 1.
Gene 2 is wrong on node 2 but gene 1 is right on node 2 as far as
congruence goes.So node 2 goes to gene 1's tree.
Is this how it goes?
>
>
>
> >>By looking at several different genes - they HAVE derived trees for
> >>many other genes you know.
>
> >And are these not more prone to HGT?I understand rRNA is highly
> >conserved and would be a bit less prone to transfer.
>
> That may be the case. But this is another point where evidence would not
> go amiss.
Consensus among genes is used to determine it?
>
>
>
> >>Actually, the article stated that 80% of 161 prokaryotes showed signs
> >>of having undergone a few HGTs, not that 80% of all prokaryote genes
> >>are HGTs.
>
> >Check out the original paper here:
> >http://www.pnas.org/content/105/29/10039
> >And read the abstract.
>
> What the paper say appears to be between those two positions. It says
> that 80% of gene families in prokaryotes have undergone horizontal
> transfer at least once - this is not the same as 80% of the genes of an
> individual prokaryote being the result of horizontal transfer. For a
> hypothetical example, consider the existence of 8 lineages, where 10% of
> the genome has been transferred from lineage A to lineage B, and so on
> up 10% being transferred from lineage H to lineage A. If there are no
> secondary transfers 80% of genes have undergone horizontal transfer, but
> only 10% of the genome of each lineage has.
Oh no,it states explicitly that it is transfers per GENOME.Not a gene
family.
>
> Note also that the paper says "Although there can be little doubt that a
> considerable component of prokaryote genome evolution over recent
> evolutionary time scales is fundamentally treelike in nature."
Oh,there's no doubt on that.The recent changes are small,the paper
does say that.But older changes will screw up higher taxa phylogeny.
>
> >BTW,here's the web of life:
> >http://en.wikipedia.org/wiki/File:Horizental-gene-transfer.jpg
>
> >>Common design doesn't explain nested hierarchy,
> >>except to the degree that it explains anything you could conceivably see
> >>and thus explains nothing.
>
> >It is unfalsifiable,no doubt about that.But just note that a web is
> >not a nested hierarchy.If over 80% of genes is transferred you can't
> >be sure what is a taxon and what it falls under anymore.
>
> Horizontal gene transfer is not the only confounding effect in attempts
> to establish deep phylogenies. It may be that we can't resolve the
> relationships between, say, the major bacterial clades. But your claim
> that we can't identify taxa is wrong.
>
> Consider a hypothetical example. Assume that we're looking at a sample
> of 100 taxa and 10 loci, and that each has been horizontally transferred
> once. (A 100% transfer rate in your terms.) The result is that in each
> case of transfer the clade that received of the gene has a different
> sister group than in the other 9 trees. The rest of the tree topology is
> congruent in all ten cases.
> --
Oh,but this is not the point at all.It is more like EACH taxa has 8
out of 10 loci transferred at least once.Major screwing up.
> alias Ernest Major
wf3h
which can be explained by chemistry. god need not apply. in fact you
creationists didn't even know the basic chemistry of the gene until it
was discovered by biochemists.
wf3h
> Replying...
>
> The urchin-chordate story has fallen apart with a little digging.The
> rest of spintronic's post seems to be slandering rather than honest
> inquiry of science methods.
adman seems to have a particular hatred for scientists and for jews.
hersheyh
And you need to read the rest of the paper. It says:
"The result indicates that on average, 81 ± 15% of the genes in each
genome were involved in LGT at some point in their history, with 61 of
the 181 individual values exceeding 90% (Table S1) and the averages
for each group given in Table 2. Once acquired, genes can be
vertically inherited within a group (39, 40), and the MLN suggests
that this has occurred for the vast majority of genes, and probably
all, given that we have inferred no LGT events from conflicting gene
trees, during prokaryote genome evolution."
The important qualification for the 81% is *at some point in their
history*. Other than that *some point*, most of the evolution of
these genes is vertical. And, given the lower frequency of LGT in
eucaryotes, we certainly cannot infer 81% for all organisms.
>
> BTW,here's the web of life:http://en.wikipedia.org/wiki/File:Horizental-gene-transfer.jpg
>
> >Common design doesn't explain nested hierarchy,
> >except to the degree that it explains anything you could conceivably see
> >and thus explains nothing.
>
> It is unfalsifiable,no doubt about that.But just note that a web is
> not a nested hierarchy.If over 80% of genes is transferred you can't
> be sure what is a taxon and what it falls under anymore.
Don't forget to add the qualifier *at some point in their history*.
Perplexed in Peoria
> "On Fri, 22 May 2009 06:10:04 -0700, in article
> <QnxRl.18880$D32....@flpi146.ffdc.sbc.com>, John Harshman stated..."
>>
>>Switch89 wrote:
>>> Here's the issue: If common descent predicts that no true chimeras may
>>> be found, how does this prediction hold up in light of our knowledge
>>> of horizontal gene transfer?
>>>
>>> If we found something with half mammalian DNA and half avian DNA,
>>> could that be chalked up to Horizontal Gene transfer? Why or why not?
>>>
>>If we found such a thing, that would require a truly bizarre
>>explanation, one that I do not so far summon up. One think it might
>>require is a world whose rules of organismal development were quite
>>different from our own. I'm reminded of the Simpsons episode in which
>>Lisa discovers a fossilized angel, and asks Steven J. Gould about it.
>>
>>Hey, how about a Cambrian rabbit with 50% bumblebee genes?
>>
>
> It wouldn't fly.
Or jump. What good is half a rabbit?
rno...@umich.edu
wrote:
> "TomS" <TomS_mem...@newsguy.com> wrote in messagenews:252998767.000...@drn.newsguy.com...
> > "On Fri, 22 May 2009 06:10:04 -0700, in article
>
> >>Switch89 wrote:
> >>> Here's the issue: If common descent predicts that no true chimeras may
> >>> be found, how does this prediction hold up in light of our knowledge
> >>> of horizontal gene transfer?
>
> >>> If we found something with half mammalian DNA and half avian DNA,
> >>> could that be chalked up to Horizontal Gene transfer? Why or why not?
>
> >>If we found such a thing, that would require a truly bizarre
> >>explanation, one that I do not so far summon up. One think it might
> >>require is a world whose rules of organismal development were quite
> >>different from our own. I'm reminded of the Simpsons episode in which
> >>Lisa discovers a fossilized angel, and asks Steven J. Gould about it.
>
> >>Hey, how about a Cambrian rabbit with 50% bumblebee genes?
>
> > It wouldn't fly.
>
> Or jump. What good is half a rabbit
For one thing, it would make a very nice cassoulet for two people.
TomS
<3a6af07c-166d-4327...@e20g2000vbc.googlegroups.com>,
rno...@umich.edu stated..."
>
>On May 22, 9:55=A0am, "Perplexed in Peoria" <jimmene...@sbcglobal.net>
>wrote:
>> "TomS" <TomS_mem...@newsguy.com> wrote in messagenews:252998767.000123be.=
>089....@drn.newsguy.com...
>> > "On Fri, 22 May 2009 06:10:04 -0700, in article
>> > <QnxRl.18880$D32.4...@flpi146.ffdc.sbc.com>, John Harshman stated..."
>>
>> >>Switch89 wrote:
>y
>> >>> be found, how does this prediction hold up in light of our knowledge
>> >>> of horizontal gene transfer?
>>
>> >>> If we found something with half mammalian DNA and half avian DNA,
>> >>> could that be chalked up to Horizontal Gene transfer? Why or why not?
>>
>> >>If we found such a thing, that would require a truly bizarre
>> >>explanation, one that I do not so far summon up. One think it might
>> >>require is a world whose rules of organismal development were quite
>> >>different from our own. I'm reminded of the Simpsons episode in which
>> >>Lisa discovers a fossilized angel, and asks Steven J. Gould about it.
>>
>> >>Hey, how about a Cambrian rabbit with 50% bumblebee genes?
>>
>> > It wouldn't fly.
>>
>
>For one thing, it would make a very nice cassoulet for two people.
>
I will resist the urge to make what could be the world's worst
pun, based on the fact that "Cambria" is Latin for "Wales".
rno...@umich.edu
> "On Fri, 22 May 2009 09:11:32 -0700 (PDT), in article
> rnor...@umich.edu stated..."
>
>
>
>
>
>
>
> >On May 22, 9:55=A0am, "Perplexed in Peoria" <jimmene...@sbcglobal.net>
> >wrote:
> >> "TomS" <TomS_mem...@newsguy.com> wrote in messagenews:252998767.000123be.=
> >> > "On Fri, 22 May 2009 06:10:04 -0700, in article
> >> > <QnxRl.18880$D32.4...@flpi146.ffdc.sbc.com>, John Harshman stated..."
>
> >> >>Switch89 wrote:
> >> >>> Here's the issue: If common descent predicts that no true chimeras ma=
> >y
> >> >>> be found, how does this prediction hold up in light of our knowledge
> >> >>> of horizontal gene transfer?
>
> >> >>> If we found something with half mammalian DNA and half avian DNA,
> >> >>> could that be chalked up to Horizontal Gene transfer? Why or why not?
>
> >> >>If we found such a thing, that would require a truly bizarre
> >> >>explanation, one that I do not so far summon up. One think it might
> >> >>require is a world whose rules of organismal development were quite
> >> >>different from our own. I'm reminded of the Simpsons episode in which
> >> >>Lisa discovers a fossilized angel, and asks Steven J. Gould about it.
>
> >> >>Hey, how about a Cambrian rabbit with 50% bumblebee genes?
>
> >> > It wouldn't fly.
>
> >> Or jump. =A0What good is half a rabbit
>
> >For one thing, it would make a very nice cassoulet for two people.
>
> I will resist the urge to make what could be the world's worst
> pun, based on the fact that "Cambria" is Latin for "Wales".
>
If you had, it would be a rare bit indeed.
Kent Paul Dolan
> Playing devil's advocate is an essential way to
> test theories.
Not really.
The accepted way in the Scientific Method to test
theories is by peer review through experiment
(perhaps replicating the original experiments) and
objective observation, not by whining that you don't
like the theory.
Doing your "devil's advocacy" by violating every
accepted convention of discourse in the area of
discussion is going to get you viewed less as a
contributor and more as an idiot.
1) Whitespace is essentially free. I haven't any
idea what botch of a text editor you're using, but
your sentences don't even have white space
separating them. To fix that, try this: put each
sentence on a separate line followed by a blank line
and see if that makes what yu type readable.
Similarly, delimit by white space each level of
quoting in what you quote, to make it easy to detect
changes of authorship.
2) Cite every claim you make to a freely retrievable
part of the professional literature. If you can't
back up a claim with such a citation, don't make it.
3) Anyone can do Blind Stupid Johnson speculation.
If you have nothing better to offer, refrain from
posting. That's not "discussion", that's "attention
whoring".
HTH
xanthian.
Unholy Joe
Alright,but when?Back in the archean?Zing!Say goodbye to
reconstructing bacterial,archean and eukaryote relationships.Or more
recently?You've got your order/class/whatever screwed.Whenever it
occured,it would have disrupted the phylogenetic relationship of the
then occuring split in evolutionary history.BTW,Francis Collins has
actually replied!Here is his reply:
-------------------------------------------------------------
Hi Jonathan,
Yes, horizontal gene transfer is a reality for prokaryotes, and makes
the definition of species rather complex. It may occasionally happen
in eukaryotes as a consequence of viral mediated gene transfer to the
germline, but there is certainly no indication that the amount of this
threatens the ability to do comparative genomics on vertebrates. I
can’t for the life of me figure out why this phenomenon would be seen
as a problem for the theory of evolution, or as a battle cry for
creationism or intelligent design.
Regards, Francis
---------------------------------------------------------------------------------
hersheyh
A different points throughout history.
> Back in the archean?Zing!Say goodbye to
> reconstructing bacterial,archean and eukaryote relationships.
At the point in time before the separation of these lineages, there
was so much horizontal transfer that all we can infer is that they all
share a common origin. Essentially all genes, even those that these
lineages share, point to a common origin before the separation of the
lineages rather than horizontal transfer after separation.
> Or more
> recently?You've got your order/class/whatever screwed.
*For that specific gene.* It is typically single genes that get
transferred in procaryotes, not half genomes. In eucaryotes, you do
get hybridization transfer before the two incipient species are
completely reproductively isolated, allowing some gene transfer. But
that just makes the speciation boundary fuzzy.
Again, because horizontal transfers are rarer than vertical changes
even in procaryotes (that is an empirical fact, not a guess) *and*,
like eucaryotes, most modes of horizontal transfer in procaryotes are
limited to spreading to *relatively* genetically close species (in
addition to being limited to species in its local environment),
identifying the rare deviation from expected vertical change is
relatively easy by seeing the *weight* of evidence from a number of
genes at any level of relationship you want.
Unholy Joe
Anyway...
Check out this:
http://www.pnas.org/content/105/29/10039/F3.large.jpg(figure from the
paper I cited)
and tell me that's a tree!Obviously we need to think on this a bit.Or
can it be that I am missing something?Can figure D in my link prove
that HGT overrides any signal by pure virtue of abundance?
Next..
The pnas paper acknowledges mosaics can be found in
prokaryotes.Thus,chimeras do exist.We do have to modify ToE because of
HGT.
Next...
>*For that specific gene.* It is typically single genes that get
>transferred in procaryotes, not half genomes. In eucaryotes, you do
>get hybridization transfer before the two incipient species are
>completely reproductively isolated, allowing some gene transfer. But
>that just makes the speciation boundary fuzzy.
So at a particular node,the HGT is light but overall it is
extreme.This means that we can still make relationships and trees?
>Again, because horizontal transfers are rarer than vertical changes
>even in procaryotes (that is an empirical fact, not a guess) *and*,
>like eucaryotes, most modes of horizontal transfer in procaryotes are
>limited to spreading to *relatively* genetically close species (in
>addition to being limited to species in its local environment),
Cite papers that support your claim?
>identifying the rare deviation from expected vertical change is
>relatively easy by seeing the *weight* of evidence from a number of
>genes at any level of relationship you want.
Even when,y'know,the deviation becomes the norm.Lemme relink:
http://www.pnas.org/content/105/29/10039/F3.large.jpg
Unholy Joe
Just because they share 80% of genes with chordates doesn't prove
they're not hybrids.These 80% they have could also be there in
urchins.And the sea squirt genes' similarity may have leaned toward
urchin or chordate depending on gene.Or perhaps it was the other 20%
that affected the outcome.
Point is,chimeras may very well exist.
Unholy Joe
>
> > Or more
> > recently?You've got your order/class/whatever screwed.
>
> *For that specific gene.* It is typically single genes that get
> transferred in procaryotes, not half genomes. In eucaryotes, you do
> get hybridization transfer before the two incipient species are
> completely reproductively isolated, allowing some gene transfer. But
> that just makes the speciation boundary fuzzy.
Uh-hum.Might I point out not a single gene but the majority of the
genome has been transferred?Or do you mean that when the bacterial
order x was forming only a few genes transferred as they have been
occuring through all 4 billion years steadily and not all at a time?Is
that what you mean?That there were a few genes being transferred when
genus/family/order/class x/y/z was forming and thus there will be only
a few anomalies for a specific genus/class/order/family but that the
anomalies, though a speck,would be omnipresent?Is that what you say?
With the added condition that although HGT has gone on slowly through
history,never deviating an individual taxon much,the very dawn of life
before the three domains was a hub of HGT even more so?
If that IS what you say,then determining ancestry is not in danger as
the concentration of HGT at a specific node will be insufficient to
tilt things much.That is a good explanation.But,if you would be so
kind as to cite a SCIENTIFIC PAPER that HGT is not localized but
gradual through the ages.If you take a look at the web of life I
posted you will note some areas that are HGT hotspots.What will you
say to that?
>
> Again, because horizontal transfers are rarer than vertical changes
> even in procaryotes (that is an empirical fact, not a guess) *and*,
> like eucaryotes, most modes of horizontal transfer in procaryotes are
> limited to spreading to *relatively* genetically close species (in
> addition to being limited to species in its local environment),
> identifying the rare deviation from expected vertical change is
> relatively easy by seeing the *weight* of evidence from a number of
> genes at any level of relationship you want.
Cite papers in support of your empirical fact.Cite papers of your
claim that HGT usually occurs in closely related species.You do
realize that HGT across a vast taxonomic distance is a well documented
phenomenon,right?
>
> > Whenever it
> > occured,it would have disrupted the phylogenetic relationship of the
> > then occuring split in evolutionary history.BTW,Francis Collins has
> > actually replied!Here is his reply:
> > -------------------------------------------------------------
> > Hi Jonathan,
>
> > Yes, horizontal gene transfer is a reality for prokaryotes, and makes
> > the definition of species rather complex. It may occasionally happen
> > in eukaryotes as a consequence of viral mediated gene transfer to the
> > germline, but there is certainly no indication that the amount of this
> > threatens the ability to do comparative genomics on vertebrates. I
> > can’t for the life of me figure out why this phenomenon would be seen
> > as a problem for the theory of evolution, or as a battle cry for
> > creationism or intelligent design.
>
> > Regards, Francis
> > ---------------------------------------------------------------------------------
John Harshman
> My posts are not showing up,What the hell?
>
> Anyway...
>
> Check out this:
> http://www.pnas.org/content/105/29/10039/F3.large.jpg(figure from the
> paper I cited)
> and tell me that's a tree!Obviously we need to think on this a bit.Or
> can it be that I am missing something?Can figure D in my link prove
> that HGT overrides any signal by pure virtue of abundance?
It's a tree. You are missing something. No, figure D can't prove that.
> Next..
>
> The pnas paper acknowledges mosaics can be found in
> prokaryotes.Thus,chimeras do exist.We do have to modify ToE because of
> HGT.
What PNAS paper? Why are you allergic to citations?
> Next...
>
>> *For that specific gene.* It is typically single genes that get
>> transferred in procaryotes, not half genomes. In eucaryotes, you do
>> get hybridization transfer before the two incipient species are
>> completely reproductively isolated, allowing some gene transfer. But
>> that just makes the speciation boundary fuzzy.
>
> So at a particular node,the HGT is light but overall it is
> extreme.This means that we can still make relationships and trees?
Yes. It all depends on the density of HGT on single branches.
>> Again, because horizontal transfers are rarer than vertical changes
>> even in procaryotes (that is an empirical fact, not a guess) *and*,
>> like eucaryotes, most modes of horizontal transfer in procaryotes are
>> limited to spreading to *relatively* genetically close species (in
>> addition to being limited to species in its local environment),
> Cite papers that support your claim?
Now that's irony.
>> identifying the rare deviation from expected vertical change is
>> relatively easy by seeing the *weight* of evidence from a number of
>> genes at any level of relationship you want.
> Even when,y'know,the deviation becomes the norm.Lemme relink:
> http://www.pnas.org/content/105/29/10039/F3.large.jpg
Look here. HGT can be treated as a variety of homoplasy. Homoplasy can
be quite extensive without obliterating phylogenetic signal. HGT is no
exception. Any HGT event alters the tree for that one gene, in quite a
simple way: by moving some subtree from one position to another. That
is, only one branch is affected. As long as there is no systematic bias
in the spacings of HGT events, this will not obscure the true tree if
enough genes are sampled.
A graphic whose meaning you seem to have misinterpreted doesn't change
that. How about a citation to the actual paper?
Unholy Joe
http://www.pnas.org/content/105/29/10039/F3.large.jpg
Keep a straight face and tell me it's a tree,and that it won't
complicate finding out ancestry of an organism.
John Harshman
> as a problem for the theory of evolution, or as a battle cry for
> creationism or intelligent design.
>
>
>
> Regards, Francis
> ---------------------------------------------------------------------------------
>
paragraph in which you criticize Howard's nearly identical reply.
The point is that if most genes do not do HGT most of the time, and if
genes do not transfer all at the same time, HGT will not screw up
phylogeny. Anomalies will stand out against the major pattern but will
not disguise it.
Unholy Joe
wrote:
> > as a problem for the theory of evolution, or as a battle cry for
> > creationism or intelligent design.
>
> > Regards, Francis
>
> I find it amazing that you could quote Collins' reply right after a
> paragraph in which you criticize Howard's nearly identical reply.
>
> The point is that if most genes do not do HGT most of the time, and if
> genes do not transfer all at the same time, HGT will not screw up
> phylogeny. Anomalies will stand out against the major pattern but will
> not disguise it.
To be fair,Collins is discussing eukaryotes here.However,you're
right.Rereading the pnas paper I found that:
1)The rate of HGT is slow,as evidenced by slow HGT in recent past
(speaking geologically presumably)
2)Most HGT events are small:
"The plotting threshold for edge weight decisively influences the
degree of connectivity among genomes that is implied in the network
graph. Only 493 ± 6 (4 ± 0.05%) edges carry 20 genes or more (Fig.
3B), 2,529 ± 17 (20 ± 0.15%) carry five genes or more (Fig. 3C),
whereas 5,773 ± 44 (47 ± 0.3%) carry only one."(pnas paper)
IOW,most HGT events occured with a few genes only.
3)The overall HGT is high with a majority of the genome having
undergone lateral transfer.However since genes undergo vertical
transfer after HGT the overall effect is staggered and phylogeny can
counter it.
But,I have just got two objections to this:
1)Take a look at the figure:http://www.pnas.org/content/105/29/10039/
F3.large.jpg
Look at figure B.,representing transfers in excess of 20 genes.Noticed
something?Yep...sometimes a thick thread of transfers representing
perhaps a thousand or even more is transferred from one point to
another.There's biased HGT for you-large,and from one single point to
another point.
2)Notice how Euryarchaeotes,Cyanobacteria,and Actinobacteria have the
highest HGT.Notice how Bacili have also had a thick thread
transferred,representing over a thousand genes perhaps,or maybe even
ten thousand.Notice that the Bacili,Cyanobacteria and Actinobacteria
have all their big HGTs(those thick threads representing over
thousands of genes)from a single node,one that laterally is in the
middle of the tree.Hmmmm..,HGT in massive quantity from a single node
to a some taxa.Importantly,in a massive quantity,from a single
node,and concentrated.Who says systemic bias is absent in HGT?
Unholy Joe
wrote:
> Unholy Joe wrote:
> > My posts are not showing up,What the hell?
>
> > Anyway...
>
> > Check out this:
> > paper I cited)
> > and tell me that's a tree!Obviously we need to think on this a bit.Or
> > can it be that I am missing something?Can figure D in my link prove
> > that HGT overrides any signal by pure virtue of abundance?
>
> It's a tree. You are missing something. No, figure D can't prove that.
canopy of vines!
>
> > Next..
>
> > The pnas paper acknowledges mosaics can be found in
> > prokaryotes.Thus,chimeras do exist.We do have to modify ToE because of
> > HGT.
>
> What PNAS paper? Why are you allergic to citations?
That pnas paper I linked to earlier.Here is it:http://www.pnas.org/
content/105/29/10039.full
>
> > Next...
>
> >> *For that specific gene.* It is typically single genes that get
> >> transferred in procaryotes, not half genomes. In eucaryotes, you do
> >> get hybridization transfer before the two incipient species are
> >> completely reproductively isolated, allowing some gene transfer. But
> >> that just makes the speciation boundary fuzzy.
>
> > So at a particular node,the HGT is light but overall it is
> > extreme.This means that we can still make relationships and trees?
>
> Yes. It all depends on the density of HGT on single branches.
Ah,I see.Thanks for clarification!The paper says so,too:
"This results in a modest proportion of recently acquired genes in
contemporary genomes, but a cumulative impact that snowballs over
evolutionary time. When all genes and genomes are considered, the tree
paradigm fits only a small minority of the genome at best"
>
> >> Again, because horizontal transfers are rarer than vertical changes
> >> even in procaryotes (that is an empirical fact, not a guess) *and*,
> >> like eucaryotes, most modes of horizontal transfer in procaryotes are
> >> limited to spreading to *relatively* genetically close species (in
> >> addition to being limited to species in its local environment),
> > Cite papers that support your claim?
>
> Now that's irony.
Ah,whatever.Maybe I am a hypocrite.
>
> >> identifying the rare deviation from expected vertical change is
> >> relatively easy by seeing the *weight* of evidence from a number of
> >> genes at any level of relationship you want.
> > Even when,y'know,the deviation becomes the norm.Lemme relink:
> >http://www.pnas.org/content/105/29/10039/F3.large.jpg
>
> Look here. HGT can be treated as a variety of homoplasy. Homoplasy can
> be quite extensive without obliterating phylogenetic signal. HGT is no
> exception. Any HGT event alters the tree for that one gene, in quite a
> simple way: by moving some subtree from one position to another. That
> is, only one branch is affected. As long as there is no systematic bias
> in the spacings of HGT events, this will not obscure the true tree if
> enough genes are sampled.
>
> A graphic whose meaning you seem to have misinterpreted doesn't change
> that. How about a citation to the actual paper?
Sure.Here's the paper:
http://www.pnas.org/content/105/29/10039.full
And yeah,you're right that relationship will not be obscured *as long
as there is no systematic bias*.Take a look at that web of life I
posted earlier.There are some locations where HGT is concentrated,thus
significantly changing the node.
The idea that enough genes sampled would reveal the true tree is
interesting,but the fact remains HGT has occured all through the
board.I wonder how that would affect nodes and relationships?
John S. Wilkins
Uhhh... you do realise that underlying the lateral transfer is a
taxonomic tree? And that this is something one is able to identify and
delineate *in order* to be able to map LGT events?
--
John S. Wilkins, Philosophy, University of Sydney
http://evolvethink.wordpress.com/
But al be that he was a philosophre,
Yet hadde he but litel gold in cofre
Unholy Joe
> Unholy Joe <transcendento...@gmail.com> wrote:
> > You should all get a nice hard stare at this from the pnas paper:
> >http://www.pnas.org/content/105/29/10039/F3.large.jpg
> > Keep a straight face and tell me it's a tree,and that it won't
> > complicate finding out ancestry of an organism.
>
> Uhhh... you do realise that underlying the lateral transfer is a
> taxonomic tree? And that this is something one is able to identify and
> delineate *in order* to be able to map LGT events?
> --
> But al be that he was a philosophre,
> Yet hadde he but litel gold in cofre
And I know that.What I want to know is how one can construct a
phylogeny at all considering the blatant HGT.Yes,they do compare
against a reference tree,but that is what I want to know:How'd they
know of the reference tree?
John Harshman
situation. Most of the time, most genes follow the same tree.
Occasionally, a gene experiences a horizontal transfer event. That
doesn't mean that gene isn't useful in estimating the tree. It merely
differs from the tree regarding a single branch.
The article says that about 80% of bacterial genes have been transferred
at least once (by the criteria applied, which as the article also says
are not foolproof). Notice that unless there was a tree, used as a
baseline, it would be impossible to say that. All that means is that you
have to look at many genes in order to get a good estimate of the tree.
This is no different in principle from any kind of homoplasy. Nearly all
sites display homoplasy, yet we are able to use those sites in
combination to estimate a single tree. Same principle here.
Ernest Major
<fb7e4227-992f-452d...@d7g2000prl.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
I haven't done more than skim the PNAS paper yet, but look at Table 2
and Figure 3B.
>Or do you mean that when the bacterial
>order x was forming only a few genes transferred as they have been
>occuring through all 4 billion years steadily and not all at a time?Is
>that what you mean?That there were a few genes being transferred when
>genus/family/order/class x/y/z was forming and thus there will be only
>a few anomalies for a specific genus/class/order/family but that the
>anomalies, though a speck,would be omnipresent?Is that what you say?
I thought that that was the general understanding of HGT in prokaryotes.
>With the added condition that although HGT has gone on slowly through
>history,never deviating an individual taxon much,the very dawn of life
>before the three domains was a hub of HGT even more so?
That's harder to tell. There are other processes that can confuse deep
phylogenies, and as far as I know we don't yet have the data to draw a
firm conclusion at this point.
>If that IS what you say,then determining ancestry is not in danger as
>the concentration of HGT at a specific node will be insufficient to
>tilt things much.That is a good explanation.But,if you would be so
>kind as to cite a SCIENTIFIC PAPER that HGT is not localized but
>gradual through the ages.
As I say, I've only skimmed it so far, but the PNAS paper that you cited
seems to support that position.
>If you take a look at the web of life I
>posted you will note some areas that are HGT hotspots.What will you
>say to that?
If you're referring to the picture you cited from WikePeida that is
rather abstract, so conclusions about hotspots would be unsafe. However
several of the 'hotspots' are instances of symbiogenesis. These don't
get in the way of inferring phylogenies.
1) acquisition of mitochondria (originally proteobacteria) by
eukaryotes.
2) acquisition of primary plastids (originally cyanobacteria) by the
ancestors of red and green algae and plants.
3) acquisition of secondary plastids by euglenoids, brown algae,
dinoflagellates, etc.
They don't get in the way because the bacterial genome is much smaller
than the nuclear genome, and the retained gene products of the bacterial
genome are targetted at the organelle.
>>
>> Again, because horizontal transfers are rarer than vertical changes
>> even in procaryotes (that is an empirical fact, not a guess) *and*,
>> like eucaryotes, most modes of horizontal transfer in procaryotes are
>> limited to spreading to *relatively* genetically close species (in
>> addition to being limited to species in its local environment),
>> identifying the rare deviation from expected vertical change is
>> relatively easy by seeing the *weight* of evidence from a number of
>> genes at any level of relationship you want.
>Cite papers in support of your empirical fact.Cite papers of your
>claim that HGT usually occurs in closely related species.You do
>realize that HGT across a vast taxonomic distance is a well documented
>phenomenon,right?
I don't know of a survey paper offhand, but I read (some of) the plant
systematics literature, and the majority of cases of HGT apply to
closely related species. Wide HGT has occurred in Amborella (which has
even picked up a moss gene) and in some parasitic plants (the intimate
contact between host and parasitic cytoplasm makes it easier for the
parasite to pick up host DNA).
HGT across "vast taxonomic distances" is indeed a well documented
phenomenon. However it remains a rare and noteworthy occurrence, while
gene transfer between closely related species is not only common, but
sufficiently easy that it is part of the regular repertoire of plant
breeders.
>>
>> > Whenever it
>> > occured,it would have disrupted the phylogenetic relationship of the
>> > then occuring split in evolutionary history.BTW,Francis Collins has
>> > actually replied!Here is his reply:
>> > -------------------------------------------------------------
>> > Hi Jonathan,
>>
>> > Yes, horizontal gene transfer is a reality for prokaryotes, and makes
>> > the definition of species rather complex. It may occasionally happen
>> > in eukaryotes as a consequence of viral mediated gene transfer to the
>> > germline, but there is certainly no indication that the amount of this
>> > threatens the ability to do comparative genomics on vertebrates. I
>> > can’t for the life of me figure out why this phenomenon would be seen
>> > as a problem for the theory of evolution, or as a battle cry for
>> > creationism or intelligent design.
>>
>> > Regards, Francis
>> >
>>
>>
>>
>>
>> >>>>>---------------------------------------------------------------------------------
>
--
alias Ernest Major
Ernest Major
<38cdd0dd-3ce6-4552...@g3g2000pra.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
>My posts are not showing up,What the hell?
>
>Anyway...
>
>Check out this:
>http://www.pnas.org/content/105/29/10039/F3.large.jpg(figure from the
>paper I cited)
>and tell me that's a tree!Obviously we need to think on this a bit.Or
>can it be that I am missing something?Can figure D in my link prove
>that HGT overrides any signal by pure virtue of abundance?
No.
>
>Next..
>
>The pnas paper acknowledges mosaics can be found in
>prokaryotes.Thus,chimeras do exist.
That appears to be a non-sequitur, or perhaps you have an idiosyncratic
definition of what is a chimaera.
The classical chimaera was an animal composed of the body of a lioness
with a tail that terminated in a snake's head and with the head of a
goat arising on her back at the center of her spine (fide wikipedia).
The term has been extended to cover other composite animals such as
sphinxes.
The word is used metaphorically to mean "an impossible of foolish
fantasy".
The term is used in biology to mean an organism composed of cells of two
different genotypes. Intraspecific chimaeras arise from the spontaneous
fusion of embryos, and interspecific chimaeras (e.g. goat-sheep
chimaeras) have been produced by experimental induction of the same.
You seem to be using the term to mean an organism with two different
genomes packed into it cells. Having searched that paper for instances
of the word mosaic I don't see any support for the existence of
chimaeras there.
>We do have to modify ToE because of
>HGT.
We did that several years ago.
>
>Next...
>
>>*For that specific gene.* It is typically single genes that get
>>transferred in procaryotes, not half genomes. In eucaryotes, you do
>>get hybridization transfer before the two incipient species are
>>completely reproductively isolated, allowing some gene transfer. But
>>that just makes the speciation boundary fuzzy.
>
>So at a particular node,the HGT is light but overall it is
>extreme.This means that we can still make relationships and trees?
The people who work in this area seem to be under the impression that
they can. It's possible that it's an illusion resulting from incomplete
data, but you seem to unreasonably convinced that they're mistaken.
>
>>Again, because horizontal transfers are rarer than vertical changes
>>even in procaryotes (that is an empirical fact, not a guess) *and*,
>>like eucaryotes, most modes of horizontal transfer in procaryotes are
>>limited to spreading to *relatively* genetically close species (in
>>addition to being limited to species in its local environment),
>Cite papers that support your claim?
>
>>identifying the rare deviation from expected vertical change is
>>relatively easy by seeing the *weight* of evidence from a number of
>>genes at any level of relationship you want.
>Even when,y'know,the deviation becomes the norm.Lemme relink:
>http://www.pnas.org/content/105/29/10039/F3.large.jpg
>
Look at figure 3B. It strikes me that that is a better representation of
what we have to work with. You've got a figure showing a few genes being
transferred here, a few there and so on. There's enough to fill in the
figure at that scale, but that doesn't mean that the thousands of
vertically transferred genes don't exist.
--
alias Ernest Major
Ernest Major
<6d4663b6-18e1-4116...@r31g2000prh.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
phylogeny. What we disagree with is your claim that it makes it nigh on
impossible.
It seems to me that that figure supports our position, not yours. Figure
3B strikes me as the best graphical representation with regard to this
question, and there's a clear tree visible in that figure. You'll note
that the figures are colour coded for the numbers of transferred genes,
and gray-coded for the size of the genome. You should also notice that
this supports what you have been told here - that horizontal gene
transfer in prokaryotes usually involves just a few genes.
--
alias Ernest Major
Ernest Major
<0cdb44c7-6f20-43a9...@w31g2000prd.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
>I just had an idea on sea squirts:
>Just because they share 80% of genes with chordates doesn't prove
>they're not hybrids.These 80% they have could also be there in
>urchins.And the sea squirt genes' similarity may have leaned toward
>urchin or chordate depending on gene.Or perhaps it was the other 20%
>that affected the outcome.
You're slow. That's what I said when you found the statement that 80% of
the genes of Ciona intestinalis were also found in vertebrates. Probably
something like 80% - but not the same 80% - will be found in
echinoderms.
>
>Point is,chimeras may very well exist.
>
The sole support so far offered for the idea that tunicates are
urchin-vertebrate hybrids is an apparently unpublished study by Michael
Syvanen.
What was written in New Scientist was
"Syvanen recently compared 2000 genes that are common to humans, frogs,
sea squirts, sea urchins, fruit flies and nematodes. In theory, he
should have been able to use the gene sequences to construct an
evolutionary tree showing the relationships between the six animals.
He failed. The problem was that different genes told contradictory
evolutionary stories. This was especially true of sea-squirt genes.
Conventionally, sea squirts - also known as tunicates - are lumped
together with frogs, humans and other vertebrates in the phylum
Chordata, but the genes were sending mixed signals. Some genes did
indeed cluster within the chordates, but others indicated that tunicates
should be placed with sea urchins, which aren't chordates. "Roughly 50
per cent of its genes have one evolutionary history and 50 per cent
another," Syvanen says."
The expected tree is (fruit
fly,nematode)(urchin,(tunicate,(frog,human)))).
The usual nematode used in these studies (Caenorhabditis elegans) is a
notorious long branch. That's going to mess up any study with such
sparse taxon sampling. I seem to recall that Drosphila melanogaster is
also a long branch taxon. Consequently I can well believe that the study
couldn't resolve the insect/nematode/deuterostome split.
With regards to the urchin/tunicate/vertebrate split, "Roughly 50 per
cent of its genes *appear to* have one evolutionary history and 50 per
cent another" (my insertion) could have another explanation other that a
chimaera. Noise in the pattern of (vertically inherited) gene changes
will blur the splits - imagine that if you have a fuzzy cloud rather
than a fork in the tree at the split. If the split is close to being a
trichotomy the two clouds will overlap. This is equivalent to a
proportion of individual gene trees not reconstructing the actual
phylogeny. I have the impression that Ciona is yet another long branch,
which exacerbates this.
So what I expect he has is a mixture of
_______________________ urchin
\_____________________ vertebrate
\__________________ tunicate
and
_______________________ vertebrate
\_____________________ urchin
\__________________ tunicate
Rather than a mixture of
_______________________ urchin
\_____________________ vertebrate
\____ tunicate
and
_______________________ urchin
\ \____ tunicate
\_____________________ vertebrate
which is what the chimaera hypothesis would predict.
We already have a few sequenced vertebrate genomes (predominantly
mammals, which doesn't help with this question). When we have a better
sampling of non-vertebrate deuterostome genomes we can test the chimaera
hypothesis more directly. (It predicts that half of the Ciona genes nest
within echinoderms, and half within vertebrates.) But we don't have, as
far as I can see, any reason to treat the hypothesis seriously.
--
alias Ernest Major
Ernest Major
<jharshman....@pacbell.net> writes
>Unholy Joe wrote:
>> My posts are not showing up,What the hell?
>> Anyway...
>> Check out this:
>> http://www.pnas.org/content/105/29/10039/F3.large.jpg(figure from the
>> paper I cited)
>> and tell me that's a tree!Obviously we need to think on this a bit.Or
>> can it be that I am missing something?Can figure D in my link prove
>> that HGT overrides any signal by pure virtue of abundance?
>
>It's a tree. You are missing something. No, figure D can't prove that.
>
>> Next..
>> The pnas paper acknowledges mosaics can be found in
>> prokaryotes.Thus,chimeras do exist.We do have to modify ToE because of
>> HGT.
>
>What PNAS paper? Why are you allergic to citations?
"http://www.pnas.org/content/105/29/10039" is a sort of citation.
Anyway, it's Dagan et al, Modular networks and cumulative impact of
lateral transfer in prokaryote genome evolution, PNAS 105(29):
10039-10044 (2008).
--
alias Ernest Major
Ernest Major
<fb7e4227-992f-452d...@d7g2000prl.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
I haven't done more than skim the PNAS paper yet, but look at Table 2
and Figure 3B.
>Or do you mean that when the bacterial
>order x was forming only a few genes transferred as they have been
>occuring through all 4 billion years steadily and not all at a time?Is
>that what you mean?That there were a few genes being transferred when
>genus/family/order/class x/y/z was forming and thus there will be only
>a few anomalies for a specific genus/class/order/family but that the
>anomalies, though a speck,would be omnipresent?Is that what you say?
I thought that that was the general understanding of HGT in prokaryotes.
>With the added condition that although HGT has gone on slowly through
>history,never deviating an individual taxon much,the very dawn of life
>before the three domains was a hub of HGT even more so?
That's harder to tell. There are other processes that can confuse deep
phylogenies, and as far as I know we don't yet have the data to draw a
firm conclusion at this point.
>If that IS what you say,then determining ancestry is not in danger as
>the concentration of HGT at a specific node will be insufficient to
>tilt things much.That is a good explanation.But,if you would be so
>kind as to cite a SCIENTIFIC PAPER that HGT is not localized but
>gradual through the ages.
As I say, I've only skimmed it so far, but the PNAS paper that you cited
seems to support that position.
>If you take a look at the web of life I
>posted you will note some areas that are HGT hotspots.What will you
>say to that?
If you're referring to the picture you cited from WikePeida that is
rather abstract, so conclusions about hotspots would be unsafe. However
several of the 'hotspots' are instances of symbiogenesis. These don't
get in the way of inferring phylogenies.
1) acquisition of mitochondria (originally proteobacteria) by
eukaryotes.
2) acquisition of primary plastids (originally cyanobacteria) by the
ancestors of red and green algae and plants.
3) acquisition of secondary plastids by euglenoids, brown algae,
dinoflagellates, etc.
They don't get in the way because the bacterial genome is much smaller
than the nuclear genome, and the retained gene products of the bacterial
genome are targetted at the organelle.
>>
>> Again, because horizontal transfers are rarer than vertical changes
>> even in procaryotes (that is an empirical fact, not a guess) *and*,
>> like eucaryotes, most modes of horizontal transfer in procaryotes are
>> limited to spreading to *relatively* genetically close species (in
>> addition to being limited to species in its local environment),
>> identifying the rare deviation from expected vertical change is
>> relatively easy by seeing the *weight* of evidence from a number of
>> genes at any level of relationship you want.
>Cite papers in support of your empirical fact.Cite papers of your
>claim that HGT usually occurs in closely related species.You do
>realize that HGT across a vast taxonomic distance is a well documented
>phenomenon,right?
I don't know of a survey paper offhand, but I read (some of) the plant
systematics literature, and the majority of cases of HGT apply to
closely related species. Wide HGT has occurred in Amborella (which has
even picked up a moss gene) and in some parasitic plants (the intimate
contact between host and parasitic cytoplasm makes it easier for the
parasite to pick up host DNA).
HGT across "vast taxonomic distances" is indeed a well documented
phenomenon. However it remains a rare and noteworthy occurrence, while
gene transfer between closely related species is not only common, but
sufficiently easy that it is part of the regular repertoire of plant
breeders.
>>
>> > Whenever it
>> > occured,it would have disrupted the phylogenetic relationship of the
>> > then occuring split in evolutionary history.BTW,Francis Collins has
>> > actually replied!Here is his reply:
>> > -------------------------------------------------------------
>> > Hi Jonathan,
>>
>> > Yes, horizontal gene transfer is a reality for prokaryotes, and makes
>> > the definition of species rather complex. It may occasionally happen
>> > in eukaryotes as a consequence of viral mediated gene transfer to the
>> > germline, but there is certainly no indication that the amount of this
>> > threatens the ability to do comparative genomics on vertebrates. I
>> > can’t for the life of me figure out why this phenomenon would be seen
>> > as a problem for the theory of evolution, or as a battle cry for
>> > creationism or intelligent design.
>>
>> > Regards, Francis
>> >
>>
>>
>>
>>
>>
>>
>>
>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>---------------------------------------------------------------------------------
>
--
alias Ernest Major
Ernest Major
<0cdb44c7-6f20-43a9...@w31g2000prd.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
>I just had an idea on sea squirts:
>Just because they share 80% of genes with chordates doesn't prove
>they're not hybrids.These 80% they have could also be there in
>urchins.And the sea squirt genes' similarity may have leaned toward
>urchin or chordate depending on gene.Or perhaps it was the other 20%
>that affected the outcome.
You're slow. That's what I said when you found the statement that 80% of
the genes of Ciona intestinalis were also found in vertebrates. Probably
something like 80% - but not the same 80% - will be found in
echinoderms.
>
>Point is,chimeras may very well exist.
>
Ernest Major
<38cdd0dd-3ce6-4552...@g3g2000pra.googlegroups.com>,
Unholy Joe <transcen...@gmail.com> writes
>My posts are not showing up,What the hell?
>
>Anyway...
>
>Check out this:
>http://www.pnas.org/content/105/29/10039/F3.large.jpg(figure from the
>paper I cited)
>and tell me that's a tree!Obviously we need to think on this a bit.Or
>can it be that I am missing something?Can figure D in my link prove
>that HGT overrides any signal by pure virtue of abundance?
No.
>
>Next..
>
>The pnas paper acknowledges mosaics can be found in
>prokaryotes.Thus,chimeras do exist.
That appears to be a non-sequitur, or perhaps you have an idiosyncratic
definition of what is a chimaera.
The classical chimaera was an animal composed of the body of a lioness
with a tail that terminated in a snake's head and with the head of a
goat arising on her back at the center of her spine (fide wikipedia).
The term has been extended to cover other composite animals such as
sphinxes.
The word is used metaphorically to mean "an impossible of foolish
fantasy".
The term is used in biology to mean an organism composed of cells of two
different genotypes. Intraspecific chimaeras arise from the spontaneous
fusion of embryos, and interspecific chimaeras (e.g. goat-sheep
chimaeras) have been produced by experimental induction of the same.
You seem to be using the term to mean an organism with two different
genomes packed into it cells. Having searched that paper for instances
of the word mosaic I don't see any support for the existence of
chimaeras there.
>We do have to modify ToE because of
>HGT.
We did that several years ago.
>
>Next...
>
>>*For that specific gene.* It is typically single genes that get
>>transferred in procaryotes, not half genomes. In eucaryotes, you do
>>get hybridization transfer before the two incipient species are
>>completely reproductively isolated, allowing some gene transfer. But
>>that just makes the speciation boundary fuzzy.
>
>So at a particular node,the HGT is light but overall it is
>extreme.This means that we can still make relationships and trees?
The people who work in this area seem to be under the impression that
they can. It's possible that it's an illusion resulting from incomplete
data, but you seem to unreasonably convinced that they're mistaken.
>
>>Again, because horizontal transfers are rarer than vertical changes
>>even in procaryotes (that is an empirical fact, not a guess) *and*,
>>like eucaryotes, most modes of horizontal transfer in procaryotes are
>>limited to spreading to *relatively* genetically close species (in
>>addition to being limited to species in its local environment),
>Cite papers that support your claim?
>
>>identifying the rare deviation from expected vertical change is
>>relatively easy by seeing the *weight* of evidence from a number of
>>genes at any level of relationship you want.
>Even when,y'know,the deviation becomes the norm.Lemme relink:
>http://www.pnas.org/content/105/29/10039/F3.large.jpg
>
John Harshman
> On May 24, 9:33 am, John Harshman <jharshman.diespam...@pacbell.net>
> wrote:
>> Unholy Joe wrote:
>>> My posts are not showing up,What the hell?
>>> Anyway...
>>> Check out this:
>>> http://www.pnas.org/content/105/29/10039/F3.large.jpg(figurefrom the
>>> paper I cited)
>>> and tell me that's a tree!Obviously we need to think on this a bit.Or
>>> can it be that I am missing something?Can figure D in my link prove
>>> that HGT overrides any signal by pure virtue of abundance?
>> It's a tree. You are missing something. No, figure D can't prove that.
> It very well is a tree,but a tree that looks totally covered by a
> canopy of vines!
The purpose of that figure is to show the vines, not to show the tree.
So of course that's what it shows.
>>> Next..
>>> The pnas paper acknowledges mosaics can be found in
>>> prokaryotes.Thus,chimeras do exist.We do have to modify ToE because of
>>> HGT.
>> What PNAS paper? Why are you allergic to citations?
> That pnas paper I linked to earlier.Here is it:http://www.pnas.org/
> content/105/29/10039.full
Thanks.
>>> Next...
>>>> *For that specific gene.* It is typically single genes that get
>>>> transferred in procaryotes, not half genomes. In eucaryotes, you do
>>>> get hybridization transfer before the two incipient species are
>>>> completely reproductively isolated, allowing some gene transfer. But
>>>> that just makes the speciation boundary fuzzy.
>>> So at a particular node,the HGT is light but overall it is
>>> extreme.This means that we can still make relationships and trees?
>> Yes. It all depends on the density of HGT on single branches.
> Ah,I see.Thanks for clarification!The paper says so,too:
> "This results in a modest proportion of recently acquired genes in
> contemporary genomes, but a cumulative impact that snowballs over
> evolutionary time. When all genes and genomes are considered, the tree
> paradigm fits only a small minority of the genome at best"
This is just another way of saying that most genes have experienced HGT
at least once.
What tells you this?
> The idea that enough genes sampled would reveal the true tree is
> interesting,but the fact remains HGT has occured all through the
> board.I wonder how that would affect nodes and relationships?
It wouldn't, as I have explained.
John S. Wilkins
> On May 24, 12:02 pm, j...@wilkins.id.au (John S. Wilkins) wrote:
> > Unholy Joe <transcendento...@gmail.com> wrote:
> > > You should all get a nice hard stare at this from the pnas paper:
> > >http://www.pnas.org/content/105/29/10039/F3.large.jpg
> > > Keep a straight face and tell me it's a tree,and that it won't
> > > complicate finding out ancestry of an organism.
> >
> > Uhhh... you do realise that underlying the lateral transfer is a
> > taxonomic tree? And that this is something one is able to identify and
> > delineate *in order* to be able to map LGT events?
>
> And I know that.What I want to know is how one can construct a
> phylogeny at all considering the blatant HGT.Yes,they do compare
> against a reference tree,but that is what I want to know:How'd they
> know of the reference tree?
The usual way - morphology, overall genetic and molecular structure
(such as Gram negative or Gram positive), and so on. Understand that
this paper is basically about single celled organisms. In the case of
what Darwin was *actually* discussing, and where the debates over tree
structure most clearly apply, there is no real problem. Animals, plants,
and fungi, the "macrobial" life of eukaryotes, form perfectly fine tree
phylogenies, lateral transfer notwithstanding.
But it's interesting that at the level of genera and up, even for single
celled organisms they are able to build a reference tree against which
to compare LGT. If "tree thinking is dead" or if "evolution is not a
tree", this should not be possible at this scale. Sure, *species* are a
moot point in microbial systematics, but still, there is a tree
structure.
It's been known for a long time - at least since Mendel - that hybridism
means that genetics do not always follow the taxonomic trees. This. Is.
Not. News.
John Harshman
>>> as a problem for the theory of evolution, or as a battle cry for
>>> creationism or intelligent design.
>>> Regards, Francis
>> I find it amazing that you could quote Collins' reply right after a
>> paragraph in which you criticize Howard's nearly identical reply.
>>
>> The point is that if most genes do not do HGT most of the time, and if
>> genes do not transfer all at the same time, HGT will not screw up
>> phylogeny. Anomalies will stand out against the major pattern but will
>> not disguise it.
>
> To be fair,Collins is discussing eukaryotes here.However,you're
> right.Rereading the pnas paper I found that:
> 1)The rate of HGT is slow,as evidenced by slow HGT in recent past
> (speaking geologically presumably)
> 2)Most HGT events are small:
> "The plotting threshold for edge weight decisively influences the
> degree of connectivity among genomes that is implied in the network
> 3B), 2,529 � 17 (20 � 0.15%) carry five genes or more (Fig. 3C),
> whereas 5,773 � 44 (47 � 0.3%) carry only one."(pnas paper)
> IOW,most HGT events occured with a few genes only.
> 3)The overall HGT is high with a majority of the genome having
> undergone lateral transfer.However since genes undergo vertical
> transfer after HGT the overall effect is staggered and phylogeny can
> counter it.
> But,I have just got two objections to this:
> 1)Take a look at the figure:http://www.pnas.org/content/105/29/10039/
> F3.large.jpg
> Look at figure B.,representing transfers in excess of 20 genes.Noticed
> something?Yep...sometimes a thick thread of transfers representing
> perhaps a thousand or even more is transferred from one point to
> another.There's biased HGT for you-large,and from one single point to
> another point.
Not true. In fact the scale only goes up to 600, there are only one or
two of those, and they are small-scale events in the sense that they
only traverse a tiny part of the tree, i.e. are within groups.
> 2)Notice how Euryarchaeotes,Cyanobacteria,and Actinobacteria have the
> highest HGT.Notice how Bacili have also had a thick thread
> transferred,representing over a thousand genes perhaps,or maybe even
> ten thousand.Notice that the Bacili,Cyanobacteria and Actinobacteria
> have all their big HGTs(those thick threads representing over
> thousands of genes)from a single node,one that laterally is in the
> middle of the tree.Hmmmm..,HGT in massive quantity from a single node
> to a some taxa.Importantly,in a massive quantity,from a single
> node,and concentrated.Who says systemic bias is absent in HGT?
>
Problems with your interpretation: The sizes of horizontal transfers are
represented not by thickness, but by color. The color scale only goes up
to 600. My copy is hard to read, but the sole thread within Bacilli that
approximates that number seems to be between two potentially related
terminal nodes for which we can't in fact rule out vertical transfer. I
don't know what "laterally is in the middle of the tree" means. Anyway,
we have a few transfers of up to a few hundred genes out of several
thousand. Very interesting, but unlikely to obscure the tree itself.
This is not systematic bias. To continue with the analogy to homoplasy,
this is a proportion of characters that have coordinated convergence,
but are still outweighed in each instance by other characters.
As the paper says, the tree is an abstraction that doesn't capture all
the history of life. But it does seem to capture most of that history,
and that portion of history can be determined despite the existence of
other histories, HGT among them.
hersheyh
> My posts are not showing up,What the hell?
I replied to this once already, and it didn't show up.
> Anyway...
>
> Check out this:http://www.pnas.org/content/105/29/10039/F3.large.jpg(figurefrom the
> paper I cited)
> and tell me that's a tree!Obviously we need to think on this a bit.Or
> can it be that I am missing something?Can figure D in my link prove
> that HGT overrides any signal by pure virtue of abundance?
>
> Next..
>
> The pnas paper acknowledges mosaics can be found in
> prokaryotes.Thus,chimeras do exist.We do have to modify ToE because of
> HGT.
>
> Next...
>
> >*For that specific gene.* It is typically single genes that get
> >transferred in procaryotes, not half genomes. In eucaryotes, you do
> >get hybridization transfer before the two incipient species are
> >completely reproductively isolated, allowing some gene transfer. But
> >that just makes the speciation boundary fuzzy.
>
> So at a particular node,the HGT is light but overall it is
> extreme.This means that we can still make relationships and trees?
>
> >Again, because horizontal transfers are rarer than vertical changes
> >even in procaryotes (that is an empirical fact, not a guess) *and*,
> >like eucaryotes, most modes of horizontal transfer in procaryotes are
> >limited to spreading to *relatively* genetically close species (in
> >addition to being limited to species in its local environment),
>
> Cite papers that support your claim?
Every paper that describes the limitations of hybridization in
eucaryotes. We know, for example, that humans are currently
reproductively isolated from even chimps and may well have been
isolated from the humans who were called neanderthals or only engaged
in a very small amount of HGT with them by hybridization. This is
quite the norm for eucaryotic hybridization. When it occurs, it tends
to be between organisms that are closely related. The chances of
humans hybridizing to a whale or dog or tunicate or sea urchin or
mollusc or insect or marigold or mushroom or paramecium is nil.
The other mode of HGT in eucaryotes, transfer by symbiosis or
infection, has occurred in the case of plasmids like mitochondria and
chloroplasts. In fact, it is likely to have occurred independently
several times historically. Viral transfer also can occur, but viruses
tend to be limited in host ranges and most eucaryotic viruses tend not
(unlike the transducing bacteriophage) have replicative mechanisms
that lead to the accidental packaging of one host's DNA and transfer
to another. They tend to only transfer copies of themselves or (and I
am thinking of certain plant viruses) transfer DNA from symbiotic
bacteria to the plant.
In bacteria, the three mechanisms of HGT are conjugation (sometimes
called bacterial sex), transduction, and transformation. Conjugation
is really a mechanism for plasmid transfer, with plasmids acting like
an internal virus that can only spread by contact. Something like an
STD. When the plasmid becomes integrated into the bacterial
chromosome, it transfers sections of the donor chromosome to the newly
"infected" recipient. This can transfer fairly large chunks of the
donor chromosome and is the only mechanism that can do so in bacteria
(eucaryotes use meiotic sex to do so). But the recognition of a
recipient is a function of a pilin protein produced by the plasmid in
the donor and its ability to recognize a surface protein on the
recipient's surface. Thus conjugation tends to only be able to engage
in the transfer process between relatively closely related bacteria.
The need to have the protein-protein interaction prevents promiscuity
of transfer.
Transduction is the passage of DNA because it accidently gets packaged
into a viral capsid. This can only transfer a limited amount of DNA
(a virus-length worth) and also tends to be limited in its
promiscuity. Like conjugation, bacteriophage tend to recognize hosts
by more or less specific protein-protein interactions. Some viruses
are more promiscuous than others and can infect a wider range of
hosts, but most tend to be able to infect only closely related
bacteria.
Transformation *in nature* also tends to transfer only relatively
small chunks of DNA. In the lab, transformation is artificially
induced by treating bacteria with chemicals or even shooting them with
nanoparticles. But in nature, only some bacteria can "naturally"
transform (and, even then, often only under certain conditions). They
do so by producing a surface enzyme that recognizes a specific DNA
sequence and then initiates a process to 'ingest' one strand
(hydrolyzing the other strand for energy). Naturally, it tends to be
closely related species that have significant number of sites with the
necessary DNA sequence. So transformation in nature too tends to be
limited to closely related species.
> >identifying the rare deviation from expected vertical change is
> >relatively easy by seeing the *weight* of evidence from a number of
> >genes at any level of relationship you want.
>
> Even when,y'know,the deviation becomes the norm.Lemme relink:http://www.pnas.org/content/105/29/10039/F3.large.jpg
Vertical transmission of genetic information (mutated or not) occurs
*every time* an organism reproduces itself. Horizontal transmission
of genetic information (the same or different) is extremely rare
relative to instances of vertical transmission, although not so rare
that it doesn't *ever* happen. On geological time frames HGT could
easily happen to produce a fixed novelty in sequence once or twice for
each gene, but the vast and overwhelming of fixed novelties in
sequence will be due to mutation and either selection or neutral
fixation of vertically transmitted genes. The latter is what produces
the consistent tree. The former is rare enough to be identifiable as
exceptions from the tree (as in the genes of mitochondria and
chloroplast, which were certainly *originally* obtained by HGT, but
have been vertically transmitted ever since).
Unholy Joe
name) swept Kolkata and promptly screwed up electricity for 48
hrs.There were riots,of course.People burnt cars in fury of
depravation of water and electricity.
Anyway..
Let me get this straight:You mean Ciona is something akin to a Chimp-
human hybrid than a lion-goat one.Fine.Doesn't change the fact it
is,in fact,a chimera,something with genes more akin to species X and
sometimes more to species Y.But then,I suppose distantly related
organisms and the word Chimera go hand in hand.Ciona perhaps no more
invalidates evolution than a liger would.
Yes,that is a good point.I'll reply to the rest later,perhaps even
after 24 hrs or some such.
Unholy Joe
> >>> as a problem for the theory of evolution, or as a battle cry for
> >>> creationism or intelligent design.
> >>> Regards, Francis
> >> I find it amazing that you could quote Collins' reply right after a
> >> paragraph in which you criticize Howard's nearly identical reply.
>
> >> The point is that if most genes do not do HGT most of the time, and if
> >> genes do not transfer all at the same time, HGT will not screw up
> >> phylogeny. Anomalies will stand out against the major pattern but will
> >> not disguise it.
>
> > To be fair,Collins is discussing eukaryotes here.However,you're
> > right.Rereading the pnas paper I found that:
> > 1)The rate of HGT is slow,as evidenced by slow HGT in recent past
> > (speaking geologically presumably)
> > 2)Most HGT events are small:
> > "The plotting threshold for edge weight decisively influences the
> > degree of connectivity among genomes that is implied in the network
> > 3B), 2,529 ± 17 (20 ± 0.15%) carry five genes or more (Fig. 3C),
> > whereas 5,773 ± 44 (47 ± 0.3%) carry only one."(pnas paper)
> > IOW,most HGT events occured with a few genes only.
> > 3)The overall HGT is high with a majority of the genome having
> > undergone lateral transfer.However since genes undergo vertical
> > transfer after HGT the overall effect is staggered and phylogeny can
> > counter it.
> > But,I have just got two objections to this:
> > 1)Take a look at the figure:http://www.pnas.org/content/105/29/10039/
> > F3.large.jpg
> > Look at figure B.,representing transfers in excess of 20 genes.Noticed
> > something?Yep...sometimes a thick thread of transfers representing
> > perhaps a thousand or even more is transferred from one point to
> > another.There's biased HGT for you-large,and from one single point to
> > another point.
>
> Not true. In fact the scale only goes up to 600, there are only one or
> two of those, and they are small-scale events in the sense that they
> only traverse a tiny part of the tree, i.e. are within groups.
"Only 493 ± 6 (4 ± 0.05%) edges carry 20 genes or more (Fig. 3B),"
So I am a bit confused as to how exactly grey lines come to be in 3B.I
thought perhaps in 3B they represented coefficients so a grey line in
3B would represent 1x20=20 genes.
What can 3B perhaps mean?!
>
> > 2)Notice how Euryarchaeotes,Cyanobacteria,and Actinobacteria have the
> > highest HGT.Notice how Bacili have also had a thick thread
> > transferred,representing over a thousand genes perhaps,or maybe even
> > ten thousand.Notice that the Bacili,Cyanobacteria and Actinobacteria
> > have all their big HGTs(those thick threads representing over
> > thousands of genes)from a single node,one that laterally is in the
> > middle of the tree.Hmmmm..,HGT in massive quantity from a single node
> > to a some taxa.Importantly,in a massive quantity,from a single
> > node,and concentrated.Who says systemic bias is absent in HGT?
>
> Problems with your interpretation: The sizes of horizontal transfers are
> represented not by thickness, but by color. The color scale only goes up
> to 600. My copy is hard to read, but the sole thread within Bacilli that
> approximates that number seems to be between two potentially related
> terminal nodes for which we can't in fact rule out vertical transfer. I
> don't know what "laterally is in the middle of the tree" means. Anyway,
> we have a few transfers of up to a few hundred genes out of several
> thousand. Very interesting, but unlikely to obscure the tree itself.
small.But they cannot mess up phylogeny too much,it seems.Thanks for
clarification of answer.Also,can you please explain this:
"Vertical edges are indicated in gray, with both the width and the
shading of the edge shown proportional to the number of inferred
vertically inherited genes along the edge."
What exactly can they mean by vertical edges?As far as I see the grey
edges connect laterally.
>
> This is not systematic bias. To continue with the analogy to homoplasy,
> this is a proportion of characters that have coordinated convergence,
> but are still outweighed in each instance by other characters.
certainly it cannot screw up phylogeny and relationships.
>
> As the paper says, the tree is an abstraction that doesn't capture all
> the history of life. But it does seem to capture most of that history,
> and that portion of history can be determined despite the existence of
> other histories, HGT among them.
to very many subjects,like HGT and tree-like phylogeny?This means
we'll have two types of biologists studying evolution now?Anyway,this
is irrelevant to the discussion.
That's all folks.I'll reply later when this message gets replied to.
rno...@umich.edu
Why do you persist in calling Ciona a hybrid/chimera or mixture of any
kind whatsoever?
Chordates and Echinoderms share membership in one major branch of the
animal world, the Deuterostomes, and would be expected to share many
genes found in their common ancestor.
You still insist on abusing the word "chimera".
John Harshman
>>>>> as a problem for the theory of evolution, or as a battle cry for
>>>>> creationism or intelligent design.
>>>>> Regards, Francis
>>>> I find it amazing that you could quote Collins' reply right after a
>>>> paragraph in which you criticize Howard's nearly identical reply.
>>>> The point is that if most genes do not do HGT most of the time, and if
>>>> genes do not transfer all at the same time, HGT will not screw up
>>>> phylogeny. Anomalies will stand out against the major pattern but will
>>>> not disguise it.
>>> To be fair,Collins is discussing eukaryotes here.However,you're
>>> right.Rereading the pnas paper I found that:
>>> 1)The rate of HGT is slow,as evidenced by slow HGT in recent past
>>> (speaking geologically presumably)
>>> 2)Most HGT events are small:
>>> "The plotting threshold for edge weight decisively influences the
>>> degree of connectivity among genomes that is implied in the network
>>> 3B), 2,529 � 17 (20 � 0.15%) carry five genes or more (Fig. 3C),
>>> whereas 5,773 � 44 (47 � 0.3%) carry only one."(pnas paper)
>>> IOW,most HGT events occured with a few genes only.
>>> 3)The overall HGT is high with a majority of the genome having
>>> undergone lateral transfer.However since genes undergo vertical
>>> transfer after HGT the overall effect is staggered and phylogeny can
>>> counter it.
>>> But,I have just got two objections to this:
>>> 1)Take a look at the figure:http://www.pnas.org/content/105/29/10039/
>>> F3.large.jpg
>>> Look at figure B.,representing transfers in excess of 20 genes.Noticed
>>> something?Yep...sometimes a thick thread of transfers representing
>>> perhaps a thousand or even more is transferred from one point to
>>> another.There's biased HGT for you-large,and from one single point to
>>> another point.
>> Not true. In fact the scale only goes up to 600, there are only one or
>> two of those, and they are small-scale events in the sense that they
>> only traverse a tiny part of the tree, i.e. are within groups.
> Actually,I am a bit confused on how to interpret it.As the paper says:
> So I am a bit confused as to how exactly grey lines come to be in 3B.I
> thought perhaps in 3B they represented coefficients so a grey line in
> 3B would represent 1x20=20 genes.
> What can 3B perhaps mean?!
Gray lines are the consensus tree, i.e. the supposed organismal
phylogeny. That's what they mean by "vertical edges". Colored lines are
putative horizontal transfer.
The vertical edges are the phylogeny. Vertical transfer as opposed to
horizontal transfer. Now there are a few odd-looking branches in that
tree -- at least two apparent triangles, which should not be in a tree.
I don't know what they intend such branches to mean, if anything.
>> This is not systematic bias. To continue with the analogy to homoplasy,
>> this is a proportion of characters that have coordinated convergence,
>> but are still outweighed in each instance by other characters.
> Small but it exists.Yet it is insignificant as you have explained,so
> certainly it cannot screw up phylogeny and relationships.
Nobody denies that there is such a thing as horiztontal transfer. The
question is whether it does screw up phylogeny. You claim, apparently,
that it does. The paper you reference doesn't support such a claim.
>> As the paper says, the tree is an abstraction that doesn't capture all
>> the history of life. But it does seem to capture most of that history,
>> and that portion of history can be determined despite the existence of
>> other histories, HGT among them.
> Why do I see the evolutionary biology division branching(pun intended)
> to very many subjects,like HGT and tree-like phylogeny?This means
> we'll have two types of biologists studying evolution now?Anyway,this
> is irrelevant to the discussion.
Not only irrelevant, but incoherent. The same person can study both
horizontal and vertical transfer. In fact, as has been pointed out, you
can't study the former without also studying the latter.