Re: Part 1 (of 3): What are major aspects of evolutionary theory?
- From: John Harshman <jharshman.diespamdie@xxxxxxxxxxx>
- Date: Sun, 08 Jan 2006 18:34:43 GMT
anon1@xxxxxxx wrote:
>>>So if I asked you: DNA-segment-duplication events are self-polarizing,
>>>because it commonly happens that a single segment of DNA gets duplicated
>>>and then the two copies gradually diverge over time, but it *never*
>>>happens that two unrelated segments of DNA gradually drift toward an
>>>exact match and then at the moment they exactly match suddenly one
>>>of the two copies is exactly/totally deleted. Agree [ ] / Disagree [ ]
>>>You'd check Disagree??
>>
>>I'd ask you what you thought you were talking about. It appears that you
>>believe we are assuming that we have an entire sequenced genome to
>>search for similar segments. That assumption is just unwarranted in
>>almost all cases.
>
>
> I agree we don't usually have whole genomes when demonstrating that
> evolution really has happened on Earth, and that common ancestry is a
> frequent theme in evolution on Earth, and that when we collect all
> known forms of life on Earth into known clades we get a very small
> number of very large clades, maybe even just one clade containing all
> Earth-life whatsoever.
>
> However I claim that even without full genomes, we have methods to
> detect duplicated genes that haven't diverged too far from each other,
> such as that method with making DNA or RNA stick together below an
> annealing temperature, whose technical name I don't know.
Hybridization. DNA-DNA hybridization (DNA hybridization for short) or
DNA-RNA hybridization. Or you could call them Southern blots or Northern
blots, depending on exactly how you were performing the trick.
> So I think my
> agree/disagree question above is reasonable, and your insistance on
> full-genome data before applying that dup-then-diverge method for
> polarizing a link/branch is unreasonable, and your complaint we don't
> yet have full-genome data is moot.
Only because you haven't thought about it very hard.
>>What you were talking about earlier were just indels that
>>you assumed were polarized in some way.
>
> In actual fact, either the event at that particular locus was a true
> insertion/duplication/immigration, or it was a true
> deletion/loss/mochikomi/emigration, it's just that we may or may not
> know which way it was at the moment, so we can or cannot decide which
> way the link/branch really went.
"Mochikomi" is not a term I know.
>>Duplications do have a direction. However, this is complicated by the
>>fact that copies can be both gained and lost,
>
>
> Like I said, if we see a pattern of two different sequences gradually
> evolving to become identical, and then suddenly at the point where they
> are exactly identical one of the two copies exactly disappears:
But we don't see this. We see various sequences with various differences
and similarities. You aren't thinking very hard about what you
hypothesize here, or about what sort of data you could observe.
> ccagagg TCAGAGCATTTCCGAATT cccgatcctatgt TCGTGGCTCGCCCTTGAA aagtgacacgacca
> ccagagg TCATAGCATGTCCGAATT cccgatcatatgt TCGTGGCTTGTCCTTGAA aagtgacacgagga
> ctagagg TCATAGCATGTCCGTAAT cctgatcatatgt TCGTAGCTTGTCCTTGAT aagggacacgagga
> ctagagg TCGTAGCATGTCCGTAAT cctgatcatatgt TCGTAGCTTGTCCTTGAT aagggatacgagga
> ctagacg TCGTAGCATGTCCTTAAT cctgatcatatgt TCGTAGCATGTCCTTAAT aagggatacgagga
> ctagacg TCGTAGCATGTCCTTAAT cctgatcatatgt aagggatacgagga
> we can be very confident we've been looking at it in reverse
> chronological order and we need to look at it the other way instead.
We don't see any of this except the terminal node sequences themselves.
Where are you getting the series from? Methinks you might be assuming a
molecular clock without noticing.
> There's just no reasonble way such a synchronization of exact
> convergent evolution and exact deletion of one of the two copies would
> occur naturally. By chance a few of the adjoining DNA bases might also
> match across the duplicated pair, and then not all the duplicate
> sequence would appear to suddenly go away:
> ccagagg TCAGAGCATTTCCGAATTCC cgatcctatgt TCGTGGCTCGCCCTTGAACC gtgacacgacca
> ccagagg TCATAGCATGTCCGAATTCC cgatcatatgt TCGTGGCTTGTCCTTGAACC gtgacacgagga
> ctagagg TCATAGCATGTCCGTAATCC tgatcatatgt TCGTAGCTTGTCCTTGATCC gggacacgagga
> ctagagg TCGTAGCATGTCCGTAATCC tgatcatatgt TCGTAGCTTGTCCTTGATCC gggatacgagga
> ctagacg TCGTAGCATGTCCTTAATCC tgatcatatgt TCGTAGCATGTCCTTAATCC gggatacgagga
> ctagacg TCGTAGCATGTCCTTAATCC tgatcatatgt CC gggatacgagga
> It would be clear in that case that all but that trailing CC part was
> duplicated in this event, the CC itself was a chance coincidental
> matching which existed prior to the duplication. Clearly we had been
> looking at this indel in backwards sequence, it was a dup, not a
> delete.
>
> In that example, we have actual DNA sequence data for each step along
> the evolutionary path.
Which makes no sense.
> In actual practice I'm assuming that we have
> homologous/aligned data from all the taxa in the current study, and
> that majority rule can be used to reconstruct much of the sequences of
> the internal nodes, so that we can clearly see the duplication event
> and the subsequent diversion, although the reconstructed internal-node
> sequences wouldn't be nice and clean as they are in the example above.
> Do you claim the reconstructed sequences are so noisy that we can't see
> the dup-then-diverge event at all as I've outlined above? If not, what
> is your quibble with the method? How can we not tell the difference
> between a duplication (whereby both of two identical copies appear in
> one node adjacent to another node with just a single copy) compared to
> a deletion (where the *only* copy is in node adjacent to node with no
> copy at all)?
By rooting the tree, or by assuming a molecular clock that roots the
tree. Or you can analyze the copies together in a single analysis, i.e.
by aligning the multiple copies to each other. That roots the part
containing the two copies. If the taxon with one copy goes within one of
the two-copy clades, that means it's a loss (and you know which copy).
If the taxon with one copy is a basal branch (not in either of the
two-copy clades), you might suppose it's an outgroup to both, but that's
not a clear inference. Sampling additional species might help solve this.
>>>What about duplications followed by later divergence of the copies?
>>
>>Plenty of that.
>
> Good. So does that mean we can locate such events along links/branches
> in the unrooted tree, and thereby determine the polarity of such
> links/branches, and thereby in each case root the portion of the tree
> to one side of that link/branch, and thereby establish most of the
> commonly-believed clades as indeed true clades?
We could, if we found the right sets of paralogous sequences for the
right taxa, and if the pattern of duplications and losses wasn't too
complex. We would do it by including homologous sequences at different
locations in a single analysis. Good words to start using here are
"paralogous" and "orthologous".
But it's hardly necessary for clades that are already well established.
> By the way, if the correct evolutionary tree has protists diversifying,
> then one clade of them becomes sponges, which diversify, and then one
> clade of sponges becomes eumetazoans, the unrooted tree should look
> like this (with true root inside AllProtists):
> AllProtists--+-----------+-----Eumetazoa (1)
> | |
> SomeSponges OtherSponges
> But I've read that perhaps the protist that yielded sponges is *not*
> the same protist that yielded Eumetazoa, that there was no path
> Protist->Sponge0>Eumetazoa, rather there were two separate paths one of
> which went directly from protist to Eumetazoa, hence the unrooted tree
> should look more like this (with true root along internal branch +--+):
> AllSponges--+-----------+-----Eumetazoa (2)
> | |
> SomeProtists OtherProtists
> Has this unrooted tree actually come out of the unrooted-tree
> construction programs, thereby validating the hypothesis that sponges
> and Eumetazoa are entirely separate clades?
I have never seen any such tree. Doesn't mean that it hasn't been
suggested before. Resolving the really deep nodes is hard.
>>> Porifera--+--------+---------Bilatera
>>> | |
>>> Placozoa +--Cnidaria
>>> |
>>> Ctenophora
>>>Do you see an unrooted tree there, or can't you see it? (Earlier you
>>> said you don't know what an unrooted tree is, despite my frequent
>>>drawing such trees for you. Well here's another.)
>>
>>You are imagining things. That is indeed an unrooted tree. It's not
>>necessarily the true unrooted tree, though. several of the branches you
>>show are contentious, as I have explained. The root probably lies within
>>Porifera.
>
> So you're opposed to the separate-protist-Sponges/Eumetazoa hypothesis,
> in favor instead of the single path protist->Sponge->Eumetazoa, i.e.
> Eumetazoa is a sub-clade of Porifera which is a sub-clade of Protist,
> not Porifera and Eumetazoa disjoint sub-clades of Protist? Among
> unrooted trees, you would favor (1) instead of (2) above?
That's what the latest information I know suggests. Your mileage may
differ, though I've never seen a tree that I can recall in which metazoa
is not monophyletic. Whether sponges are monophyletic is the question.
> Or did you mean to say that the root could either be within Porifera or
> between Porifera and the nearest internal node, which would be
> consistent with either the single-path or the dual-path hypothesis
> respectively, and we'd know which of the two based on what happens to
> the tree when you include some protists too and thereby narrow down
> where the root of that five-animal-taxa tree should be placed?
No idea. Your sentences are sometimes too long to parse well. But we're
arguing here about topology, not rooting. I haven't seen a tree that you
could split so as to have some protists and some metazoans on both
sides. Maybe you have.
>>>Now the big question: Per the latest data, have I drawn the correct
>>>unrooted tree connecting those five taxa? If not, which unrooted tree
>>>would you draw to connect those same five taxa?
>>
>>It's unclear. All three possible resolutions of the
>>Bilateria/Cnidaria/Ctenophora trichotomy are currently in contention.
>
>
> Then I guess I need to re-draw the unrooted tree with an unresolved node:
> Porifera--+--------+---------Bilatera (3)
> | |\
> Placozoa | Cnidaria
> |
> Ctenophora
> Now if I were to include protists too, with disagreement as to the
> single-path (protist->sponge->Eumetazoa) or dual-path
> (protist#1->sponge, protist#2->Eumetazoa) hypothesis, what a mess I'd
> have, especially since Eumetazoa doesn't include Placozoa, and I don't
> know whether they fit in the clade with Eumetazoa, i.e. both Placozoa
> and Eumetazoa from protist#2, or whether they have yet a third path
> from protist#3->Placozoa? Do you have a preference?
My preference, based on what I've seen recently, would be for a
paraphyletic Porifera. Don't know about Placozoa.
>>>>... it's unclear whether they are a clade or whether one or the other
>>>>is closer to Bilateria than the other.
>>>
>>>All I'm asking about here is the **unrooted** tree. With an unrooted
>>>tree, "closer to" is meaningless, so please don't use such nonsense
>>>wording in this part of the discussion.
>>
>>OK. It's unclear if they form a possible bipartition of the unrooted
>>tree. Happy?
>
> Yes, although the word "bipartition" would seem to properly apply only
> to the *pair* of halves, such as Gib+Ora/Chi+Gor+Hum, not to just a
> single half by itself such as just Gib+Ora. Checking on Google...
I accept your quibble.
[snip belaboring of the point]
>>>So anyway, how would you draw the unrooted tree of just the five animal
>>>taxa, not including the one protist taxon, based on the latest
>>>information? The way I drew it 43 lines earlier, or some other way?
>>
>>I would reduce the resolution.
>
> What do you mean?? If you don't recognize any modern names for any of
> the small clusters of taxa, then you have to include all five in the
> tree. Do you mean you would combine two nodes into a single unresolved
> node, as I did at (3) above?
Yes.
>>Radiata is actually an old name for a number of groups that also
>>included echinoderms.
>
>
> (Warning, in the following I'm talking about rooted trees:)
> Oh. My source showed Radiata as including only Cnidaria and Ctenophora.
> Echinodermata was within Deuterostomes which was within Coelomates
> which was within Bilateria which was sister to Radiata. So I don't see
> that as a valid reason to abandon Radiata as I saw it defined. But
> that's moot if the node Eumetazoa = Bilateria+Cnidaria+Ctenophora is
> unresolved as you pointed out, so for that reason I'd accept abandoning
> that old taxon until and unless the node finally resolves into
> Bilateria/(Cnidaria+Ctenophora) in which case I'd like to revive the
> old taxon Radiata per the definition (Cnidaria+Ctenophora) I saw.
>
> It would be *really* interesting if Cnidaria and/or Ctenophora turned
> out not to be a side branch but to actually include Bilateria within
> its clade. Given we're talking about the ancestry of human beings,
> I think it's important to completely sequence some examples of
> each phylum and compare with known Bilateria to resolve this node.
Agreed, except that complete sequencing would be wasted effort here.
What you need is enough sequence of the right type. For a given amount
of effort, it's probably much better to get, say, 50,000 bases for 20
species than 250,000 bases for 4 species.
> (Warning, back to unrooted trees here:)
> For example we might get:
> AllProtists--+---------+------------+--AllBilateria
> | | |
> AllCtenophora SomeCnidaria MoreCnidaria
> (Here I've left out sponges, in dispute, and Placozoa, my laziness.)
> .
>
.
- Follow-Ups:
- References:
- Prev by Date: Re: Is Pat Robertson the dumbest human to ever live?
- Next by Date: Re: Utah anti-evolution bill filed
- Previous by thread: Re: Part 1 (of 3): What are major aspects of evolutionary theory?
- Next by thread: Re: Part 1 (of 3): What are major aspects of evolutionary theory?
- Index(es):
Relevant Pages
|
Loading
