More shotgunning of ocean microbial ecology.
Perplexed in Peoria
microbe genomes is reported in the current (1/27/06) Science.
Among the interesting results is an 'epidemic' of 'jumping genes'
or introns in very deep-water samples.
Here is a press release:
http://www.nsf.gov/news/news_summ.jsp?cntn_id=105769
My opinion: This technique is going to reveal a lot of
new and exciting stuff about microbial systematics, and
perhaps about OOL over the next decade. To say nothing
of the ecological info about global material cycles.
It is too bad that our ability to culture and characterize
these things in terms of morphology and metabolism
is not keeping pace with our ability to discover and
even sequence them.
an...@sci.sci
Venter's current project, right?
> My opinion: This technique is going to reveal a lot of
> new and exciting stuff about microbial systematics, and
> perhaps about OOL over the next decade. To say nothing
> of the ecological info about global material cycles.
I agree. I think it's a wonderful use for his original
shotgun-sequencing idea, far beyond its original use.
> It is too bad that our ability to culture and characterize
> these things in terms of morphology and metabolism
> is not keeping pace with our ability to discover and
> even sequence them.
No, you have the logic backwards. In the past, we looked for our keys
only under the streetlight where it was easy. We put a sample of
natural stuff in a petri dish with some standard nutrient (agar,
straw-water, etc.), and 90% of the species of micro-organisms promptly
died because they don't like that nutrient, and 90% of the remaining
died out because the oxygen concentration was wrong, and 90% of the
remaining died out because the temperature was wrong, etc., so only a
very tiny fraction of micro-organism species could be cultured, and we
had no idea what else there had been in the original sample, like the
many keys not under the lampposts that are completely invisible.
But with Venter's whole-ecosystem shotgun-sequencing, we'll have
accurate statistics of the DNA sequences in the raw sample, including
the 99.9% that didn't culture easily. We can then pick some common
sequence that never appears in any previously-cultured species, and
devise a probe for that particular sequence. We can then use that probe
to detect which new samples have that same sequence, and then subject
such samples to various environments (nutrient, oxygen, temperature,
etc.) and test which factors keep the individuals with that DNA
sequence alive the longest, and thereby hill-climb toward an
environment where some species containing that DNA sequence will
actually survive and replicate and grow in culture. Once we've achieved
that, we can use traditional sequencing (shotgun etc.) to learn its
full sequence, thereby both culturing and full-sequencing a species
that we never could culture before.
In summary, using Venter's new methodology plus specific-sequence
probes will actually help solve the problem of the vast majority of
life not previously cultured.
One possible speed-up technique: Devise a probe that easily penetrates
into cells and flouresces when bound to the specific DNA sequence, but
only if the cell is still alive, because it requires ATP or some other
living-cell biochemistry to trigger the binding or the flourescence.
Then establish a gradient of nutrient or oxygen or temperature and
watch for the band of flourescence that shows the narrow range of that
factor that keeps the cells alive the longest. For example, with random
environment, the cells might die in seconds, but with optimum oxygen
level they might last minutes, then with both oxygen and temperature
optimum they might last tens of minutes, then with all that plus the
correct nutrient they might last hours which is long enough to
reproduce.
The nice thing about this whole methodology is that it can be
parallelized easily. You shotgun-sequence some regions of ocean, and
pick the hundred most-common DNA sequences not known from any
already-sequenced organism. Then in parallel you fabricate one hundred
probes, for the corresponding hundred common sequences, all using the
same basic technology. Then you run 100 petri dishes in parallel, one
with each probe, and independently optimize oxygen concentration and
temperature and nutrient etc. for each of the different petri dishes.
(Actually petri dishes may be obsolete, there's some new micro-channel
device for running lots of parallel chemical or microbial experiments,
all under computer control, right?) If you find two different sequences
with identical optimum environmental conditions, you suspect those
sequences co-reside within a single species, so you can then work with
whichever of them has best signal/noise ratio and put all the other
presumed copies of the same species on back burner. This might reduce
your work by a factor of 2 or 3, avoiding the mistake of going to a lot
of trouble of sequencing the same genome in two separate projects at
the same time. Once you have one of the genomes fully sequenced, a
simple computer algoirthm can check whether the other sequences really
do co-reside within this one genome, thereby validating your decision
to put them on back burner earlier. If you find one of the back-burner
sequences *not* appearing in the complete genome of the front-burner
species for the same exact optimum environment, *then* you re-activate
that specific one, thereby finding a second species that happens to
enjoy the exact same environental conditions, which might indicate some
symbiosis between the two whereby they track each other (or one tracks
the other, such as preditor/prey) in adapting to the environment.
.
Perplexed in Peoria
<an...@sci.sci> wrote in message news:drque3$1q4k$1...@darwin.ediacara.org...
> > A preliminary report from another shotgun sampling of ocean
> > microbe genomes ...
>
> Venter's current project, right?
Gee, I don't know. Venter certainly pioneered the field, but my
impression was that this report was from an independent group.
Thanks for the detailed analysis. Yes, I realize that shotgunning
opens up a lot of new approaches for culturing. Whereas it used to
be that the culturing horse dragged the sequencing cart, these days
it may make more sense to have the cart leading the horse.
My only regret is that the horse has (so far) failed to keep up with
the new rapid pace of progress. There is a lot that can only be
learned from relatively pure cultures - morphologies, the distribution
of lipid types in membranes, etc. In fact, it is my impression that
until some kind of 'culturing' is accomplished to improve the signal/noise
ratio, it is probably not going to be possible to patch all of those
contigs together into genomes.
But I am just an amateur OOL enthusiast observing all of this activity
from the sidelines and cheering the participants on. The thing that I
find most exciting in all of this is that it might eventually reveal
something pretty surprising - like non-cellular lifeforms. Unfortunately,
the new techniques don't seem to help much in discovering what I would
really like to see discovered, which is a lifeform that is cellular
but not based on DNA. Your 'key under the lightpole' analogy works in
the opposite direction here. We are discovering new kinds of DNA-based
life, because the technology for manipulating DNA has recently taken
a huge leap forward. It is as if the city has just turned on the
street lights. Exciting, but it doesn't help if the key we are looking
for doesn't happen to be under one of the lamps.
IRR
<an...@sci.sci> wrote in message news:drque3$1q4k$1...@darwin.ediacara.org...
>> microbe genomes ...
>
> Venter's current project, right?
>
>> My opinion: This technique is going to reveal a lot of
>> new and exciting stuff about microbial systematics, and
>> perhaps about OOL over the next decade. To say nothing
>> of the ecological info about global material cycles.
>
> I agree. I think it's a wonderful use for his original
> shotgun-sequencing idea, far beyond its original use.
>
>> It is too bad that our ability to culture and characterize
>> these things in terms of morphology and metabolism
>> is not keeping pace with our ability to discover and
>> even sequence them.
>
> No, you have the logic backwards. In the past, we looked for our keys
> only under the streetlight where it was easy. We put a sample of
Certainly some wonderful potential applications, but I think PiP's original
point is that you can do some record shattering amount of shotgun
sequencing, and then dive into the remarkable types of investigations as you
describe -- learn exactly where a gene is present within a certain
environmental niche, see that it is upregulated under conditions A, B, and
C, and downregulated under conditions X, Y, and Z -- and still know nothing
concrete about the actual function of the protein that the gene codes for.
OoLers are all about mechanism :).
William Morse
$1...@darwin.ediacara.org:
Apparently the press release got the date wrong - I can't find it in the
1/27/06 Science. I'll try to keep an eye out for it.
Yours,
Bill Morse
Perplexed in Peoria
"William Morse" <wdm...@twcny.rr.com> wrote in message news:ds86aa$1e0b$1...@darwin.ediacara.org...
> "Perplexed in Peoria" <jimme...@sbcglobal.net> wrote in news:drlnbg$2eev
> $1...@darwin.ediacara.org:
>
> > A preliminary report from another shotgun sampling of ocean
>
> Apparently the press release got the date wrong - I can't find it in the
> 1/27/06 Science. I'll try to keep an eye out for it.
It is there. At least in the online TOC.
Community Genomics Among Stratified Microbial Assemblages in the Ocean's Interior
Science 27 January 2006: 496-503
an...@sci.sci
> > > microbe genomes ...
> > Venter's current project, right?
> Gee, I don't know. Venter certainly pioneered the field, but my
> impression was that this report was from an independent group.
This is the first I've heard of anybody except Venter's group using
this methodology. It seems to be either good or bad news depending on
how you look at it. It's good that other people like his idea and are
trying it too, and that brings in additional funding not available to
Venter alone. But it would be bad if the different groups each kept
their data confidential (except for published summaries/analysis), and
failed to exchange it between the groups, because it's only by merging
all available data that the signal/noise ratio can be optimized.
Re my idea of using shotgun statistics to locate the most common
unknown sequences and specifically probe for them and culture whatever
cells happen to match:
> Thanks for the detailed analysis. Yes, I realize that shotgunning
Glad you liked my ideas.
> There is a lot that can only be learned from relatively pure cultures
> - morphologies, the distribution of lipid types in membranes, etc.
Yes, and response to various environmental factors by up/down
regulating expression of various genes too, thereby giving a clue what
the purpose of a particular gene and its protein product or other
product might be. The shotguns are like a master index of what might be
learned, then the pure cultures are a tool for actually learning.
Without the index you don't know what might be profitable to clone (you
don't know what's common enough that you have a chance of finding it
again), but without the followup work what was the point of the index?
> In fact, it is my impression that until some kind of 'culturing' is
> accomplished to improve the signal/noise ratio, it is probably not
> going to be possible to patch all of those contigs together into
> genomes.
Ah, you treat it as s/n ratio: Given a single genome, with only a tiny
bit of contamination from other genomes, you see a clear signal of the
one high-frequency sequence (whole genome), whereby you can patch
together the correct sequence by simply connecting all the
equally-high-frequency contigs, ignoring any others, which have an
order of magnitude lower frequency. But with a mix of many genomes, you
can't find the contigs you want among all the chatter from others,
because the few you want to use don't have much higher frequency than
the majority you don't want. That's only a slightly different way of
looking at the problem compared to the hypothetical example I posted a
few weeks ago:
http://groups.google.com/group/talk.origins/msg/53429b34bd38f489
= Message-ID: <a0b04$43b2f2dc$c690c02a$23...@TSOFT.COM>
Date: Wed, 28 Dec 2005 12:13:07 -0800
Skip to this text (a couple typos corrected here):
! ... What if two genomes diverged from a common
! ancestor, with a highly conserved segment between two moderately
! evolved segments, and none of the single shotgun reads spans the entire
! distance between the two evolving regions because the conserved region
! is longer than any single shotgun read?
and then follow the discussion, including this diagram:
! ACCGCCCCCATCCCCTCCCTC AGCCCAGCCAAACGACACTAA
! \ /
! AAGTCCTATCCAATCTACTGTACTTTGCCCCTAC
! / \
! GTTGTTATTACCTCCCTTTCA TATTCTTTCGCCTAGTGATTA
Note in this case we don't have a target genome in mind, we are simply
trying to use *all* the whole-ecosystem shotgun reads to reassemble as
many complete genomes as there's enough data for. But we're frustrated
by examples like that, where two different genomes, of similar
frequencies, have a shared/conserved segment between two sets of
different sequences, with no clue which on the left matches which on
the right. If one genome were *much* more common than the other, we
could simply assume the most common left-sequence matches the most
common right-sequence, then the remaining pair also match in the
less-common genome. But without at least an order of magnitude
difference in frequency, noise in sampling could easily swamp the
signal, making it unclear which matches which across the shared middle.
More recent discussions have covered true-junk DNA (totally
non-constrained, freely drifting) and not-junk DNA (constrained, highly
conserved, even if it doesn't code for any protein or RNA and we can't
think of any other use it might have). Given that exons are conserved
while introns and other stuff are often true-junk (not all, but most),
we may actually find a common pattern of conserved segment of
significant size (longer than a single shotgun read) sandwiches between
two non-conserved sequences, in closely-related species of eukaryotic
microbiota (true protists). But in Prokaryotes, we might not have this
problem so often, because they simply don't have a lot of junk DNA in
the first place, so virtually everything is highly conserved, so we
won't find lots of cases of great disparity between degree of
conservation along a stretch of DNA. More likely we'd find a pattern of
long stretches of nearly identical DNA because those genes are
essential and haven't needed to change anywhere within a single small
clade, with occasional genes that have changed a lot because sister
taxa within the small clade have recently adapted to different
environmental conditions or different ecological niches within the same
habitat, one of them colonizing a new niche while the other staying in
the ancestral niche. WIth a pattern of:
DivergentGene-Allelle#1
/ \
veryLongConservedSequence moreConservedSequence
\ /
DivergentGene-Allelle#2
we wouldn't have the problem of guessing which combination of two
divergent genes is the correct match, since there's only one such
divergent sequence in the whole genome. So reconstructing either
complete genome is trivial. If you want to clone them, and don't care
which, use the conserved sequence as a probe. If you want to pick which
of the two to clone, use an allele of the divergent gene as a probe.
But what if adaption to a new niche required two (2) different genes to
change, all the rest of the genome remaining fixed by stable selection,
and the two genes that diverged aren't nearby each other:
DG1-A#1 DG2-A#1
/ \ / \
stable veryLongConservedSequence stable
\ / \ /
DG1-A#2 DG2-A#2
Now you have a really bad case of not being able to span the middle by
a single shotgun read. There are two possible solutions:
- Use any one of the four divergent allelles (two allelles each of two
genes) as a probe and clone that particular species, and sequence it,
then whatever remains must be the correct genome for the other species.
- Or if you sample lots of environments and notice that the relative
frequencies of the allelles track up/down as pairs, for example DG1-A#1
is more common in shotgun reads in the same water samples where DG2-A#1
is more common, then you can guess that those two are parts of the same
genome that has adapted to one kind of environmental conditions, and
the alleles which don't track those two are the other genome which has
adapted to another kind of environmental conditions. (Even if the two
species have adapted to different niches within the same general
environment, still that niche may be more common in some parts of the
ocean than in others, causing one species to be more common in some
samples than in others.) This of course requires you retain all the
original data, so after pooling the data to maximize total overlap of
shotgun reads to allow maximal splicing together in easy cases, you can
then go back to the original data sets to perform this
correlation-of-frequency analysis.
- Make four probes, for the four allelles (2 each of 2 genes), each a
different color of flourescence (hint: Red, Green, Blue, and either
near-UV or near-IR, or just use narrow-band visible-light filters for
multi-spectral image-taking). Then do just the first stages of
hill-climbing toward cloning, such as x-axis and y-axis two-dimensional
gradients for oxygen and temperature for example, and see which two of
the four color-flourescence signals are correlated across those two
gradients. For example, if R=G1A1 G=G1A2 B=G2A1 and V=G2A2, and if your
image looks like this:
oxygenated
RVR V
VRRV R
cold V RGV hot
VB G BGGB
BG BGGB
reducing
Then you know that R and V (G1A1 and G2A2) probably belong to one
single genome, while G and B (G1A2 and G2A1) probably belong to the
other single genome. If you get a more complicated pattern, such as
three clusters of phosphorescence, with R+V in one cluster, R+B in a
second cluster, and G+V in the third cluster, then you can guess you
have three different species that share the same two allelles of the
two genes, possibly via horizontal gene flow, or successive adaption
where one of the two genes split before the other.
Hmm, maybe this 2-d multi-spectral-phosphorescence cluster-experiment
actually should be performed before any fullfledged cloning effort,
just to get an idea how many genomes we're dealing with, so we can make
sure to tease out each of them separately and not get two of them
cloned together which would be totally confusing. Of course if we have
only a single probe because we're trying to clone the species that
matches only a single common contig, then 2-d clustering will show all
clusters of the same color, and it'll be harder to guess whether we're
dealing with two actual clusters due to two different species that have
the same contig, or whether random noise is making one cluster seem to
split into two. But even in that case, 2-d gradients would seem to give
better information than just trying one gradient at a time.
> But I am just an amateur OOL enthusiast observing all of this activity
> from the sidelines and cheering the participants on.
But OOL and evolution are two very different topics, with only a slight
bit of overlap. This shotgun sequencing is going to help with
evolution, not with OOL, so why are you following this thread?
> The thing that I find most exciting in all of this is that it might
> eventually reveal something pretty surprising - like non-cellular
> lifeforms.
But still based on DNA, like some viruses? Hmm, I wonder whether
Venter's procedure selects only DNA, not RNA, thereby ignoring the
genomes of most of the viruses?
> Unfortunately, the new techniques don't seem to help much in
> discovering what I would really like to see discovered, which is a
> lifeform that is cellular but not based on DNA.
On Earth? Any such totally different cellular life form would either
have been driven extinct, or would be so common we would have noticed
it already. There'd be severe direct competition between DNA life and
non-DNA life, and I just can't see how non-DNA life could survive at
very low frequency unless it were in close symbiosis with DNA life, in
which case we would have already observed it under the electron
microscope when studying cultures of whatever bacteria it lives with.
We'd see a few cells that looked very different from our
bacteria-culture cells, yet kept appearing in those cultures again and
again. Somebody would have noticed them and reported them, and then
somebody would have investiaged them. Remember how bacteriophages
showed up like sore thumbs when examining bacteria under ordinary light
microscope a century ago, leading to the discovery of the entire class
of viruses. If bacteriophages hadn't existed, then we wouldn't have
discovered viruses until the electron microscope was invented, when we
could finally discover ordinary viruses which are smaller. Also notice
that now we've indeed discovered nanobacteria recently. We've also
discovered tiny bubbles of bilipids that appear in some of our pure
chemical products. Anything else that size that was symbiotic with
ordinary DNA life, we would have seen by now, including non-DNA cells.
If and when we probe Europa's under-ice ocean, *then* we'll have to
take a whole new fresh look to see what kinds of life are there.
RMA-world life, or pre-prokaryotes, or auto-catalytic non-cellular
replicators, etc., who knows what we might find there, not yet having
evolved to what we'd consider life as we know it, because it's so cold
there that replicator-chemistry runs slowly and doesn't evolve as
quickly as life on Earth did. Maybe chemistry in Europa's ocean runs
1000 times slower than chemistry in Earth's ocean, so in 4.5 billion
years of time they've gotten only the equivalent of 4.5 million years
of Earth's early biochemistry, which hasn't yet reached life like we
see on Earth 3.5 billion years ago (several hundred million years after
the late heavy bombardment ended).
> Your 'key under the lightpole' analogy works in the opposite
> direction here. We are discovering new kinds of DNA-based life,
> because the technology for manipulating DNA has recently taken a huge
> leap forward. It is as if the city has just turned on the street
> lights. Exciting, but it doesn't help if the key we are looking for
> doesn't happen to be under one of the lamps.
We had good reason to suspect that a lot of genomes existed in DNA
cellular life but hadn't been cultured nor sequenced nor even generally
characterized yet. Venter's preliminary results showed not only was
that suspicion correct but it was a major understatement as to our past
lack of knowledge of such life.
We currently have no reason to suspect that any non-DNA life exists
anywhere on Earth, unless you count really primitive chemical systems
as if they were life. Pasteur's argument that any new primitive life
would promptly be eaten by existing highly-developed, is a good reason
to expect the total lack of any life except that which branches off
from the three main clades of life. With 99% of those three huge clades
gone extinct due to competition with other species of the same three
huge clades, it's just not reasonable to expect that somehow a totally
undiscovered kind of life would evade 99% extinction. So either it's
really really big so that after 99% is gone there's still something
left, in which case we would have noticed it by now, or it wasn't
really big, and after 99% reduction it drifted into total extinction.
Someday we may the ability to account for every single molecule in a
sample of ocean water or other environment, whereby we will be able to
flag any single molecule that can't be explained by known mechanisms of
inorganic chemistry or biochemistry. Single molecules all totally
different from each other and each within range of tornado in junkyard,
would indeed be interpreted as just tornado in junkyard, but if a whole
bunch of similar molecules are flagged, then somebody would investigate
where they could have come from, and what their subsequent course of
chemistry should be as predicted by chemical theory. At that time,
maybe we'll discover your unknown form of life. Or maybe there simply
is no such on Earth today, as I believe will turn out to be the case.
.
an...@sci.sci
> then dive into the remarkable types of investigations as you describe
> -- learn exactly where a gene is present within a certain environmental
> niche, see that it is upregulated under conditions A, B, and C, and
> downregulated under conditions X, Y, and Z -- and still know nothing
> concrete about the actual function of the protein that the gene codes
> for.
There are two factors here, selection of one genome in one environment
and another genome in another environment via differential survival and
replication of the two species, and moment-by-moment regulation of
expression of genes within a single species due to moment-by-moment
changes in the immediate environment. The former you get directly from
the shotgun sequencing itself. (You notice that one genome is found
mostly in the shade under corals while the other genome is found mostly
on the top surface of the open ocean, for example.) But the latter
requires additional work. You need to pick which apparent genes you
wish to study, calculate (via genetic code) what polypeptide they will
produce, devise a probe for that polypeptide sequence, then go back to
the RealWorld and actually probe for that polypeptide sequence under
different circumstances. The same method might work for genes that code
for RNA instead of polypeptides. And yet, as you point out, you still
don't really know why a particular polypeptide gets expressed whenever
the bacteria is exposed to bright light and not expressed when it's in
the shade, for example. Yes, we agree, and thanks for pointing out that
possibility I hadn't mentionned myself.
But my original point was in rebuttal to the lament that all these
other studies are lagging behind the raw shotgun sequencing. The other
poster's point seemed to be that shotgun sequencing pulls way ahead,
leaving all other studies behind. My counter point is that shotgun
sequencing pulls way ahead for the moment, but quickly allows probes
for specific sequences to greatly enhance capabilities in all these
other areas (culturing specific species which then yields complete
genomes and enables studies of protein function, tracking adaption to
local environments, tracking up/down regulation in response to
moment-by-moment changes in environment), so instead of lamenting the
way everything else falls behind the lead horse, we should be
applauding how the lead horse drags all the other horses forward with
it even if behind it relatively. It's like lamenting over poverty
nowadays, compared to long ago when there wasn't such a disparity
between rich and poor, when in fact we should be applauding how even
the poorest today have a better life then even the richest did during
the stone age.
.
Perplexed in Peoria
<an...@sci.sci> wrote in message news:dsek65$16lf$1...@darwin.ediacara.org...
> > > > A preliminary report from another shotgun sampling of ocean
> > > > microbe genomes ...
> > > Venter's current project, right?
> > Gee, I don't know. Venter certainly pioneered the field, but my
> > impression was that this report was from an independent group.
>
> This is the first I've heard of anybody except Venter's group using
> this methodology.
If you Google this group with the search string
shotgun ocean
you will find some previous postings by me and others on this subject.
In the one titled "Oceans viewed through a shotgun", you will find
a link to an excellent review article on the field.
Larry Moran
an...@sci.sci <an...@sci.sci> wrote:
>> > > A preliminary report from another shotgun sampling of ocean
>> > > microbe genomes ...
>> > Venter's current project, right?
>> Gee, I don't know. Venter certainly pioneered the field, but my
>> impression was that this report was from an independent group.
>
> This is the first I've heard of anybody except Venter's group using
> this methodology. It seems to be either good or bad news depending on
> how you look at it. It's good that other people like his idea and are
> trying it too, and that brings in additional funding not available to
> Venter alone. But it would be bad if the different groups each kept
> their data confidential (except for published summaries/analysis), and
> failed to exchange it between the groups, because it's only by merging
> all available data that the signal/noise ratio can be optimized.
The field is now known as metagenomics. It was begun by Norman Pace back
in 1984. He analyzed many microbial communities including hydrothermal
vents, Yellowstone hot springs, and marine environments. Check out his list
of publications at ...
http://pacelab.colorado.edu/Publications/publications.html
The pioneering papers were published between 1984 and 1990. Look at the
following papers: #61. #64, #65, #69, #74, #76, #80, #88, #91, #92,
#105, #106, and #111. By the early 90's there were several groups
doing similar projects including some who were looking at intestinal
fauna. I don't Venter go into the field until after 2000.
Here's a good review of the perils and pitfalls.
Green Tringe, S. and Rubin, E.M. (2005) Metagenomics: DNA
Sequencing of Environmental Samples.
Nature Reviews Genetics 6: 805-814.
All of the sequence data must be deposited in public sequence databases.
There are several attempts to collect this data in metadatabases that
integrate the sequence information with details about the environment.
Here's one example ...
Larry Moran
Perplexed in Peoria
<an...@sci.sci> wrote in message news:dsek65$16lf$1...@darwin.ediacara.org...
> > PiP wrote:
> > But I am just an amateur OOL enthusiast observing all of this activity
> > from the sidelines and cheering the participants on.
>
> But OOL and evolution are two very different topics, with only a slight
> bit of overlap. This shotgun sequencing is going to help with
> evolution, not with OOL, so why are you following this thread?
Because until we trace evolution back to the point where it started,
we will have no idea what kind of organism we need to generate from
our hypothetical OOL processes.
You can learn more about this viewpoint of mine by finding a thread
in this group about a year ago entitled something like "OOL I -
Manifesto and Metatheory".
> > The thing that I find most exciting in all of this is that it might
> > eventually reveal something pretty surprising - like non-cellular
> > lifeforms.
>
> But still based on DNA, like some viruses? Hmm, I wonder whether
> Venter's procedure selects only DNA, not RNA, thereby ignoring the
> genomes of most of the viruses?
As I understand it, it selects whatever is amplified by the PCR
enzymes. We THINK that that means DNA, but it is possible that
we may be in for a surprise.
> > Unfortunately, the new techniques don't seem to help much in
> > discovering what I would really like to see discovered, which is a
> > lifeform that is cellular but not based on DNA.
>
> On Earth? Any such totally different cellular life form would either
> have been driven extinct, or would be so common we would have noticed
> it already.
For a proponent of, and enthusiast for, an exploration technique like shotgun
sampling, you seem curiously certain about what is and is not 'out there'.
I think that there may be many surprises in store for us. One thing
that has only been learned within the past 10 years is that there is
a whole kingdom-level branch of the tree of life - the crenarchaeota -
which we didn't even know existed and which we still don't know much
about. But it may turn out to represent more biomass than the rest
of Gaia put together. So I'm not sure you are right that if it is
common we would have noticed already.
William Morse
Thanks - and also thanks to Wirt Atmar, who e-mailed me with the
information. I tried thanking him by e-mail but Maxwell's demon
apparently didn't like the look of the electrons - the message was
undelivered. Yes the article is there - I am apparently just suffering
from selective blindness.
An interesting approach to studying microbial ecology. Given the
difficulty of culturing many microbes, especially those which require
high pressure, it may be the only way to do an accurate survey of deep
water populations.
One of the interesting findings was that the deep-water microbes had more
genes for antibiotic synthesis. The authors propose that this is linked
to a "greater role for a surface-attached life style", which makes sense
given that the available food sources at depth will often be
particulates. But I would not have guessed beforehand that deep-water
microbes could be a good place to look for new antibiotics.
Yours,
Bill Morse
Perplexed in Peoria
"William Morse" <wdm...@twcny.rr.com> wrote in message news:dsug7b$1bo4$1...@darwin.ediacara.org...
> One of the interesting findings was that the deep-water microbes had more
> genes for antibiotic synthesis. The authors propose that this is linked
> to a "greater role for a surface-attached life style", which makes sense
> given that the available food sources at depth will often be
> particulates. But I would not have guessed beforehand that deep-water
> microbes could be a good place to look for new antibiotics.
Hmmm. If these are novel 'antibiotics', then I wonder how they know
that they serve an antibiotic function. Guess I will have to make a
trip to the library and read the article.
William Morse
No, the sequences they found with antibiotic function aren't themselves
novel. I was simply speculating that an ecological community that is
mostly unstudied but shows an elevated level of antibiotics would be a
good place to search for new antibiotics. Sorry if I misrepresented the
article.
As an aside, one of the other types of genes found in higher levels in
the deep-water microbes were those for polysaccharides - also associated
with a surface attached lifestyle. So even if nobody finds a new
antibiotic, we might find a better version of xanthan gum :-)
Yours,
Bill Morse
Perplexed in Peoria
"William Morse" <wdm...@twcny.rr.com> wrote in message news:dt0ma4$2bbn$1...@darwin.ediacara.org...
> "Perplexed in Peoria" <jimme...@sbcglobal.net> wrote in
> news:dt00a4$2002$1...@darwin.ediacara.org:
>
> >
> > "William Morse" <wdm...@twcny.rr.com> wrote in message
> > news:dsug7b$1bo4$1...@darwin.ediacara.org...
> >> One of the interesting findings was that the deep-water microbes had
> >> more genes for antibiotic synthesis. The authors propose that this is
> >> linked to a "greater role for a surface-attached life style", which
> >> makes sense given that the available food sources at depth will often
> >> be particulates. But I would not have guessed beforehand that
> >> deep-water microbes could be a good place to look for new
> >> antibiotics.
> >
> > Hmmm. If these are novel 'antibiotics', then I wonder how they know
> > that they serve an antibiotic function. Guess I will have to make a
> > trip to the library and read the article.
>
> No, the sequences they found with antibiotic function aren't themselves
> novel. I was simply speculating that an ecological community that is
> mostly unstudied but shows an elevated level of antibiotics would be a
> good place to search for new antibiotics. Sorry if I misrepresented the
> article.
Your apology is too late. I have already made the trip to the library
and read the article. ;-)
> As an aside, one of the other types of genes found in higher levels in
> the deep-water microbes were those for polysaccharides - also associated
> with a surface attached lifestyle. So even if nobody finds a new
> antibiotic, we might find a better version of xanthan gum :-)
Great! We either gain a cure for deep-water diseases or an elastic
suitable for high pressure wet suits. If both, we have somewhere
to go to after climate change and overpopulation make surface living
unattractive.
an...@sci.sci
> > > from the sidelines and cheering the participants on.
> > But OOL and evolution are two very different topics, with only a slight
> > bit of overlap. This shotgun sequencing is going to help with
> > evolution, not with OOL, so why are you following this thread?
> Because until we trace evolution back to the point where it started,
> we will have no idea what kind of organism we need to generate from
> our hypothetical OOL processes.
IMO you have chosen a futile course. We may be able to estimate the
LUCA(s) with reasonable accuracy, but earlier is mostly speciulation.
Without a target to work back toward, there's no way to refine that
speculation to be more reasonable. Extrapolation (off the end from all
known data) is always less accurate than interpolation (between two
known points of data). I respectfully propose that you work from the
other end for a while, doing lab experiments to see what kind of
abiogenesis might happen under different circumstances, and seeing how
far forward you can get along various paths. Then compare the paths
forward from abiogenesis with the estimates backward from LUCA(s), and
see if any pairs are a good match, where you can fill in the middle
plausibly. If you can get *any* reasonble polymeric linear digital
genome from abiogenesis, then perhaps you can devise a series of
takeovers to fill the gap between there and the LUCA(s).
I mentionned a week or so ago that I now have what I consider good
reason to believe the "RNA world" was the last genomic system prior to
our current DNA world, and I even proposed a sequence of reasonable
events that constituted the RNA-to-DNA takeover. Did you see that?
<http://groups.google.com/group/sci.bio.evolution/msg/384023d6ee400f24>
= Message-ID: <dt86sa$2u0h$1...@darwin.ediacara.org>
> You can learn more about this viewpoint of mine by finding a thread
> in this group about a year ago entitled something like "OOL I -
> Manifesto and Metatheory".
I see you present some additinal reasons why "RNA world" likely came
before our current DNA world, such as co-enzymes. But that whole thread
seems to be only the very first part of your thesis. I looked for
the other parts, but all I could find were:
OOL X - The origin of the RNA world.
OOL XIII - The "Zymes"
OOL XI - Elements of a Lipid-World Model.
Are there any parts II or III or IV or V or VI or VII or VIII or IX?
Hmm, in my search for OOL only the four above came up (three shown, and
part I Manifesto). But if I search manually for each set of roman
numerals individually, I get for example:
OOL II - Building Blocks
OOL III - Connecting the Blocks
It'll take me some time to find all the pieces and study them to see
whether you ever came up with my idea of why RNA genome came before DNA
genome based on current transcription to RNA and then genetic code
applied to RNA rather than DNA. (Or you could cut to the chase and tell
me whether you did or did not include any such argument in your
multi-thread thesis, and maybe even which of the threads has that
particular tidbit. Also, did you specify a sequence of steps involved
in the RNA-to-DNA takeover, as I did? Even if the sequence turns out to
be wrong, I like *some* proposed sequence to consider, rather than just
vagueness. There *is* a difference between a thought-out plausable
sequence, compared to a "just so" story whose only appeal is to
children and Kipling fans willing to ignore all consideration of
plausability. I rather like my idea that synchronization between RNA
replication and cell division was keeping DNA operations as only a
backup systme, until synchronization was changed to DNA-cell at which
point suddenly the roles between backup and mainsystem reversed. Maybe
there was an intermediate state of *no* tight synchronization, somewhat
like a pre-prokaryote version of multi-nucleated Fungi cells, or maybe
an intermediate state of double synchronization competing with each
other, and finally DNA-cell synch beat out RNA-cell synch.)
Hmm, now that my mind is on this subject, here's a brand new thinking
of the most likely sequence:
- Longstanding "RNA world" includes: synch between RNA replication and
cell division, plus RNA-to-amino-acid genetic code and RNA enzymes.
Then the following new mechanisms during the "takeover":
- Basic machinery to make DNA units by modifying the already-existing
RNA units. Somehow one of the four didn't go the same way as the other
three, hence the Uracil being lost and replaced by whatever. But since
DNA is more stable than RNA, this allowed longterm storage of surplus
RNA units just as fat provides longterm storage of fatty acids and food
calories.
- Enzyme to polymerise those loose DNA units to make strands for
mechanical support. Sequence and length were both random, requiring
careful tuning to avoid runaway length consuming all available
resources, and strength was very unreliable.
- RNA-to-DNA transcription, so that length of RNA genome directly
controlled both length and sequence of DNA. At this point, "junk" RNA
in ur-karyotes was good because then the sequence used to make nice
strong DNA strands of the appropriate lengths didn't need to match the
sequence used to code amino-acid strands for proteins. Since there was
just one strand of RNA in the genome, copies had to pass from the
genome part of the cell to the other part where they were chopped at
weak points to make the various-lengths of aa-exons and DNA-exons.
(Multiple RNA strands would have made synch between RNA replication and
cell fission impossible, so that's why I dismiss that possibility.)
- DNA-to-RNA transcription, so that now DNA serves both a structural
function (nice strong strands) and a RNA-backup function. Now *all*
RNA, not just the strand-RNA, can get transcribed to DNA, for different
purposes. Only the exons get transcribed back to RNA, when restoring
from backup.
- DNA replication, not synchronized yet, just "backup of backup" for
additional redundancy, tricky regulation to avoid runaway consumption
of resources.
- Some synch between cell division and DNA replication, thereby
providing just the right number of copies of backup, for best tradeoff
between redundancy and consumption of resources.
- Mutations in DNA backup cause conflicts between the RNA true genome
and the DNA backup genome. If the DNA mutation was bad, that clade goes
extinct, and the mutation is erased. If the DNA mutation was *good*,
the conflict remains until by chance the only copy of the RNA conflict
is deleted and restoration from the DNA backup eliminates the conflict.
Because DNA is more stable than RNA, hence fewer mutations in DNA, and
because most non-neutral mutations are harmful, there's selection
pressure to disable RNA replication entirely.
- One day that mutation indeed happens, and that clade enjoys selective
advantage over all clades that still have the conflicting DNA+RNA
replicating system, so now the system with only DNA replicase takes
over, with horizontal gene flow converting some other clades and not
others, the not-converted clades all going extinct. Hello DNA World!!
So did you have any sequence like that for the RNA-to-DNA takeover?
.
an...@sci.sci
threads, and forgot to read the rest of your message, so I stopped my
followup prematurely. Now I'm back at your recent article:
> > But still based on DNA, like some viruses? Hmm, I wonder whether
> > Venter's procedure selects only DNA, not RNA, thereby ignoring the
> > genomes of most of the viruses?
> As I understand it, it selects whatever is amplified by the PCR
> enzymes. We THINK that that means DNA, but it is possible that
> we may be in for a surprise.
I would think that if there's anything in the ocean samples other than
DNA that gets amplified, it would have shown up by now. But maybe
Venter et al mistakenly thought it was contamination and ignored it.
Are you willing to state odds for/against something (from the ocean)
other than DNA being amplified by the enzymes that Venter et al are
using? Is there a sort of scientific-futures market where different
people can place wagers at various odds to see a sort of concensus
prediction? I thought I saw something about that once. Google shows:
<http://hanson.gmu.edu/ifwired.html> (just an article)
<http://www.richel.org/resume/Betting_on_the_future.htm> (another article,
which also suggests betting replace peer review for publishing)
<http://hanson.gmu.edu/ideafutures.html> (another article)
<http://www.ideosphere.com/> (Ah, the real thing, except their search
form for claims is broken, no submit button!)
> > > Unfortunately, the new techniques don't seem to help much in
> > > discovering what I would really like to see discovered, which is a
> > > lifeform that is cellular but not based on DNA.
> > On Earth? Any such totally different cellular life form would either
> > have been driven extinct, or would be so common we would have noticed
> > it already.
> For a proponent of, and enthusiast for, an exploration technique like
> shotgun sampling, you seem curiously certain about what is and is not
> 'out there'.
I'm not certain. I'm just expressing my estimate of the odds.
> I think that there may be many surprises in store for us.
Well, of course, there are surprises every month, nearly every week.
Who would expect in January that a journalist who was spending her
energy getting at the truth of what's going on in Iraq, opposing the
lies of the USA and other powers, would be kidnapped by the anti-USA
forces and held for random? They should have tried to recruit her, not
ransom her.
But a non-DNA cellular form of life in abundant numbers in the ocean
waters? Nah, not likely. Lots of surprises, but probably not that one.
> One thing that has only been learned within the past 10 years is
> that there is a whole kingdom-level branch of the tree of life - the
> crenarchaeota
Um, kingdom?
<http://tolweb.org/tree?group=Crenarchaeota>
<http://www.ebi.ac.uk/interpro/DisplayIproEntry?ac=IPR000196>
<http://www.peripatus.gen.nz/Taxa/Archaea.html>
<http://aem.asm.org/cgi/content/full/64/11/4333>
or only phylum?
<http://en.wikipedia.org/wiki/Crenarchaeota>
<http://sn2000.taxonomy.nl/Taxonomicon/TaxonTree.aspx?id=108997>
<http://merops.sanger.ac.uk/cgi-bin/getmibytaxon?level=Phylum&taxon=Crenarchaeota&type=P>
<http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=525246>
I think we should abandon the traditional classification names and use
a purely cladistic naming system, where no taxon has any specific
"rank" except within sexually reproducing eukaryotes where "species"
has a precise meaning, and perhaps my new definition of "genus" I'll
post sometime in the next few days when I have a big block of time to
write it up properly. It'll be a followup to:
Message-ID: <qGLGf.28741$F_3....@newssvr29.news.prodigy.net>
so watch for it to appear!
> which we didn't even know existed and which we still don't know much
> about.
Well members of it were discovered in 1977, but indeed until we
sequenced some members we didn't know how very different from other
archaebacteria they are, and until we did sequencing of whatever we
found in various places we didn't realize how widespread they were. If
we consider prokaryotes only, what fraction of the TOL would they
represent, not in terms of current diversity, but in terms of splitting
of the TOL exactly in half at each binary branch, or splitting more
than two equal ways at the very top where there's no outgroup and no
obvious intrinsically-polarized branches/links so we can't pin down the
correct place for the root of the tree?
> But it may turn out to represent more biomass than the rest of Gaia
> put together.
There ain't no Gaia. That's a modern myth started by Lovelock and Gould
and Margulis. I think they all went a little batty.
But anyway, what's your current estimate of fraction or percentage of
biomass that is Archaebacteria (sometimes renamed Archaea), and the
fraction that is specifically Crenarchaeota?
> So I'm not sure you are right that if it is common we would have
> noticed already.
We certainly noticed at least one species of Crenarchaeota as early as
1977. It uses DNA, just like all other cellular life, so it didn't
raise flags until we sequenced it and compared it to other genomes. If
it had used something other than DNA for its genome, it would have
immediately made big news. No non-DNA cellular life has been found in
all the time from 1977 to now, unless there's something your lab is
keeping secret. (Did you watch NBC's "Surface"?)
.
Perplexed in Peoria
<an...@sci.sci> wrote in message news:du610h$19e$1...@darwin.ediacara.org...
> Oops, I got distracted by trying to find all the pieces of your OOL
> threads, and forgot to read the rest of your message, so I stopped my
> followup prematurely. Now I'm back at your recent article:
> > > But still based on DNA, like some viruses? Hmm, I wonder whether
> > > Venter's procedure selects only DNA, not RNA, thereby ignoring the
> > > genomes of most of the viruses?
> > As I understand it, it selects whatever is amplified by the PCR
> > enzymes. We THINK that that means DNA, but it is possible that
> > we may be in for a surprise.
>
> I would think that if there's anything in the ocean samples other than
> DNA that gets amplified, it would have shown up by now. But maybe
> Venter et al mistakenly thought it was contamination and ignored it.
> Are you willing to state odds for/against something (from the ocean)
> other than DNA being amplified by the enzymes that Venter et al are
> using?
Hmmm. After thinking about it, I guess I would have to say that the
odds are strongly against my suggestion. I was about to suggest
that there might be slight deviations from canonical DNA - something
like dU instead of dT or perhaps a methylated dC - but when I stop
to think that the source material is being chopped by restriction
enzymes before it is amplified, even that seems very unlikely.
> > > > Unfortunately, the new techniques don't seem to help much in
> > > > discovering what I would really like to see discovered, which is a
> > > > lifeform that is cellular but not based on DNA.
> > > On Earth? Any such totally different cellular life form would either
> > > have been driven extinct, or would be so common we would have noticed
> > > it already.
> > For a proponent of, and enthusiast for, an exploration technique like
> > shotgun sampling, you seem curiously certain about what is and is not
> > 'out there'.
>
> I'm not certain. I'm just expressing my estimate of the odds.
>
> > I think that there may be many surprises in store for us.
>
> Well, of course, there are surprises every month, nearly every week.
> Who would expect in January that a journalist who was spending her
> energy getting at the truth of what's going on in Iraq, opposing the
> lies of the USA and other powers, would be kidnapped by the anti-USA
> forces and held for random?
Someone who stops to think that this is exactly the kind of journalist
who ventures outside the green zone. But we are straying far OT.
> But a non-DNA cellular form of life in abundant numbers in the ocean
> waters? Nah, not likely. Lots of surprises, but probably not that one.
>
> > One thing that has only been learned within the past 10 years is
> > that there is a whole kingdom-level branch of the tree of life - the
> > crenarchaeota which we didn't even know existed and which we still
> > don't know much about.
>
> > But it may turn out to represent more biomass than the rest of Gaia
> > put together.
>
> ... what's your current estimate of fraction or percentage of
> biomass that is Archaebacteria (sometimes renamed Archaea), and the
> fraction that is specifically Crenarchaeota?
Well, I did say 'may ... represent more biomass'. For it to do so, it
will have to be discovered that (1) the biota of the deep crust is more
significant than most people think (20% chance) and (2) most of that
biota is Crenarchaeota (perhaps 50% chance). That answers the question
you probably should have asked.
Perplexed in Peoria
<an...@sci.sci> wrote in message news:du610e$15k$1...@darwin.ediacara.org...
> > > > But I am just an amateur OOL enthusiast observing all of this activity
> > > > from the sidelines and cheering the participants on.
> > > But OOL and evolution are two very different topics, with only a slight
> > > bit of overlap. This shotgun sequencing is going to help with
> > > evolution, not with OOL, so why are you following this thread?
> > Because until we trace evolution back to the point where it started,
> > we will have no idea what kind of organism we need to generate from
> > our hypothetical OOL processes.
>
> IMO you have chosen a futile course. We may be able to estimate the
> LUCA(s) with reasonable accuracy, but earlier is mostly speciulation.
> Without a target to work back toward, there's no way to refine that
> speculation to be more reasonable. Extrapolation (off the end from all
> known data) is always less accurate than interpolation (between two
> known points of data). I respectfully propose that you work from the
> other end for a while, doing lab experiments to see what kind of
> abiogenesis might happen under different circumstances, and seeing how
Tim Tyler expressed much the same opinion back a year ago. However,
I would claim that 'working forward' has led nowhere - producing
a string of negative results - except that the negative results
(like a 1% yield of glycine and racemic alanine in the Miller experiment)
get trumpeted as positive results. By contrast, to the extent that
our grasp of OOL is better now than it was in Oparin's day, it is
the result of 'working backward' (like the RNA world hypothesis).
> I mentionned a week or so ago that I now have what I consider good
> reason to believe the "RNA world" was the last genomic system prior to
> our current DNA world, and I even proposed a sequence of reasonable
> events that constituted the RNA-to-DNA takeover. Did you see that?
> <http://groups.google.com/group/sci.bio.evolution/msg/384023d6ee400f24>
> = Message-ID: <dt86sa$2u0h$1...@darwin.ediacara.org>
Yes, I saw it.
> > You can learn more about this viewpoint of mine by finding a thread
> > in this group about a year ago entitled something like "OOL I -
> > Manifesto and Metatheory".
>
> I see you present some additinal reasons why "RNA world" likely came
> before our current DNA world, such as co-enzymes. But that whole thread
> seems to be only the very first part of your thesis. I looked for
> the other parts, but all I could find were:
> OOL X - The origin of the RNA world.
> OOL XIII - The "Zymes"
> OOL XI - Elements of a Lipid-World Model.
> Are there any parts II or III or IV or V or VI or VII or VIII or IX?
Yes. But the series petered out at XIII
> Hmm, in my search for OOL only the four above came up (three shown, and
> part I Manifesto). But if I search manually for each set of roman
> numerals individually, I get for example:
> OOL II - Building Blocks
> OOL III - Connecting the Blocks
> It'll take me some time to find all the pieces and study them to see
> whether you ever came up with my idea of why RNA genome came before DNA
> genome based on current transcription to RNA and then genetic code
> applied to RNA rather than DNA. (Or you could cut to the chase and tell
> me whether you did or did not include any such argument in your
> multi-thread thesis, and maybe even which of the threads has that
> particular tidbit. Also, did you specify a sequence of steps involved
> in the RNA-to-DNA takeover, as I did?
No, I didn't speculate on the RNA->DNA transition. In fact, I suggested
(outside the main argument) that there really isn't strong evidence
that DNA did not come first.