## Tuesday, April 16, 2013 ... /////

### Panspermia follows from Moore's law

Just several years ago, when this blog already existed, I got converted to the idea of panspermia which says that the earliest forms of life were born somewhere in outer space before the Earth was formed – and they just found Earth to be a particularly hospitable destination where they could further evolve and flourish.

Now, the Physics arXiv Blog discusses the preprint

Life Before Earth
by two biologists, Alexei Sharov and Richard Gordon, who aren't affiliated with any "top theoretical universities" that brings us a cool new argument in favor of the panspermia paradigm.

The argument is based on a computer science meme, Moore's law, that claims that the number of transistors on a circuit is exponentially increasing with the doubling time that is either 18 or 24 months or so.

Just for the sake of completeness, this is how the number of transistors grew from the 1970s:

Growing by a factor of 1 million (20 doublings) in 40 years translates almost exactly to a 2-year doubling time.

I think you will agree that the line – with a logarithmic y-axis – is indeed remarkably straight. The case of biology is designed to be as similar as possible. Instead of transistors, they count the functional genome's base pairs (bp). The result looks like this:

The functional non-redundant genome size apparently grows 10-fold in a billion of years or so.

In particular, the red dots are ordered truly linearly – the growth is exponential – and there's no doubt that if you extrapolate the number back to "the origin of Earth" 4.7 billion years ago, you still get a genome size above 10,000 base pairs.

The most obvious criticism you may raise is that this naive straight-line "Moore's law" just fails. The line is allowed to bend. Equivalently, the explosion of the genetic information in an animal could have grown exponentially but with a much shorter doubling time when life was really getting started. Maybe. However, note that you could raise the same objection in the case of the integrated circuits but Moore's law seemed OK even when computers were very, very young.

Even when Moore's laws in this particular form – one could of course guess that other precise quantities grow exponentially but this "transistor/base pair count" seems to be particularly well-behaved – are correct, a question is whether we may "prove" or at least find an argument that "explains" why the doubling time should be constant. I don't have such a "theoretical proof" and I am interested in it if you know one.

However, as I have previously stated, there is a reason why I think that the idea that "all steps to life had to occur on Earth" is probably wrong. It's just unnecessary and based on a one-size-fits-all egalitarian reasoning that has obvious limitations.

If you believe that even short DNA molecules were first born on Earth, you probably defend such a viewpoint by pointing out that our blue, not green planet is so unusually hospitable and obeys the conditions we need for life. So it had to be here.

This claim seems to treat the word "life" in a very sloppy, ambiguous way. It's the higher life similar to ours that needs the oceans to swim in and other achievements that Earth seems to boast and employ to defeat many competing tourist destinations. However, when we talk about the origin of life, we're not talking about the origin of mammals. The latter could have taken place on Earth but that does not imply that the former event had to occur terrestrially, too.

In fact, I find it pretty obvious that the concentration of the habitats to friendly hot spots is becoming more focused as one goes towards the higher life forms. If a professional (=paid) string theorist is picked as the prototype of the highest life form on Earth, its or his or her habitat is confined to several floors at Princeton, New Jersey, Cambridge, Massachusetts, and a dozen of comparably small places that are getting less and less characteristic. These are much more special places than the places where mosquitos and fungi flourish – which is almost everywhere on Earth.

If you revert this "increasing concentration of genetic capital" to reconstruct the distant past, it seems reasonable to think that the "optimum place" for some extremely old, primordial life forms was less localized than the Earth's surface. After all, it's probably not true that these life forms needed a nice amount of planetary gravity – gravity becomes almost irrelevant for very small cells etc. much like it is irrelevant in particle accelerators. Similarly for water etc.

I think that the volume of "cavities in rocks" and similar things that are flying around all stars in the Milky Way is probably higher than the volume right above the Earth's surface where we expected terrestrial life. In particular, even if you insist that the temperature is right, you don't need an Earth-like planet at the right radius that orbits a given star. The rocks may be just enough.

I don't want to be excessively specific about the "less demanding conditions" that the early life could have depended upon because I haven't spent too much time with thinking about all such possible conditions. But I surely do believe the general principle that the primitive life is less choosy and less picky than advanced life. This very general principle is enough for me to prefer the hypothesis that the very early life was born at rather generic extraterrestrial places and it just found the Earth to be a particularly hospitable place for a further evolution.

What do you think?

#### snail feedback (51) :

I think it's very reasonable to consider the panspermia hypothesis as a possibility. Creationists sometimes try to show that is extremely unlikely that life arose from non-life (abiogenesis) here on Earth, and although I'm not sure their calculations make much sense considering the many unknown factors, adding extraterrestrial possibilities makes their argument weaker still. Increasing the size of the "laboratory" must increase the chances of abiogenesis happening randomly, and adding diversity to the environmental conditions adds possibilities of it happening in ways it could never have happened on Earth.

For a theoretical argument, how about the rate of mutations per time unit scales with the length of the genome and a mutation can alter the length of the genome?

I'm a bit wary of projecting exponential trends in general and think five data points is a bit little, but the basic argument that some simple measures of genetic information grow exponentially seems quite intuitive.

If someone's wondering how to think of "genetic complexity" as used in the paper, here's an article that shoots down some of the more intuitive interpretations.

http://www.nature.com/scitable/topicpage/eukaryotic-genome-complexity-437

This measure has little correlation with what we'd think of as "higher life". The main point is that it's a number that seems to grow exponentially, regardless of how you interpret it.

Right, JollyJoker, what grows exponentially is a number, not some emotional non-quantitative interpretations, and that's how it should be. The punch line here is that if the number of useful base pairs was much much higher than 1 immediately at the beginning, it's pretty unlikely that it was born here without a pre-history. It doesn't matter whether the organisms we are actually talking about look attractive, smart, or advanced to us.

The mutation rate is changing with species i.e. time, see e.g. the human numbers here
http://www.genetics.org/content/156/1/297.full

but that may just mean that some other quantities we could invent aren't increasing exponentially with a fixed doubling time.

I suppose that the mutation rate was higher at the beginning. It needs to be lower for advanced forms because most of the functioning has to be kept nearly constant. Advanced life forms are already "optimized enough" which means that it's desirable to suppress mutations, left-wing revolutions, and similar things that would mostly send us away from the near-optimum that has already been found. Am I wrong?

I have often thought panspermia is credible. Why not? My doubts about an Earth origin of life is that the oldest forms of life are still bacteria which conists of DNA and a cell membrane. It would seem to me the earliest life form would be a relative of ribosomal RNA and that's it. There is some RNA based life forms without a membrane but they are rare. Second, the early Earth does not seem to have the right chemistry to generate life. Third, if rocks can be jettisoned from one planet to another (for example, Mars to Earth) why couldn't some bacterial form of life that can be dormant for any number of years be created on some other planet, even outside are solar system, hitch a ride on a comet of interstellar type of meteor and land on Earth. I think a good place to look for where life might have started would be Titan. I believe it was warmer in the distant past.

If “our” life did originate on earth it seems probable that life originated elsewhere at an earlier time irrespective of Moore’s-Law arguments. Certainly ours was not the first habitable planet in our Galaxy. It now seems that there are some 200 billion planets having suitable temperatures in this galaxy and a significant fraction of those surely cooled down to Goldilocks temperatures a few billion years before earth did.
Either a ridiculously improbable miracle happened here on earth or there is a hell of a lot of life out there.
The possible answers to Fermi’s conundrum are narrowing, aren’t they?

I'd probably argue against the terms "optimized enough" and "near-optimum". (I haven't thought this through much but the idea that we're near perfect exactly now doesn't sit well with me.

Otherwise i have no disagreement. Yes, the mutation rate is slowed down quite a lot (about a factor of 1/1000) in the parts of the human genome that matter. I assume simpler organisms had no such mechanisms or less effective ones.

"Human mitochondrial DNA has been estimated to have mutation rates of ~3× or ~2.7×10−5 per base per 20 year generation (depending on the method of estimation);[6] these rates are considered to be significantly higher than rates of human genomic mutation at ~2.5×10−8 per base per generation."

http://en.wikipedia.org/wiki/Mutation_rate#Variation_in_mutation_rates

The time when prokaryotes and eukaryotes first evolved is not known. "There is no consensus among biologists concerning the position of the eukaryotes in the overall scheme of cell evolution. Current opinions on the origin and position of eukaryotes span a broad spectrum including the views that eukaryotes arose first in evolution and that prokaryotes descend from them, that eukaryotes arose contemporaneously with eubacteria and archeabacteria and hence represent a primary line of descent of equal age and rank as the prokaryotes, that eukaryotes arose through a symbiotic event entailing an endosymbiotic origin of the nucleus, that eukaryotes arose without endosymbiosis, and that eukaryotes arose through a symbiotic event entailing a simultaneous endosymbiotic origin of the flagellum and the nucleus, in addition to many other models, which have been reviewed and summarized elsewhere." reference in wikipedia prokaryote entry.

Besides prokaryotes are not an uniform group and the time bacteria and archea first evolved certainly differs. Finally there is no reason to think that bacteria and archea, when they first appeared on Earth, were particularly similar to modern day species - 3.5 billion years is a long time.

Well, the anthropic principle can be applied to this. Even if intelligent life is currently emerging all over the galaxy, wouldn't it still be unlikely that we see no one who evolved 10M years before us?

It is exactly what I have thinked. But it has a little difference

Right. Well, the possibility that life could have evolved at many places before the Earth already weakens the need for a creator.

On the other hand, if true, it does show that this particular criticism by the creationists - that the initial life was still way too complex to evolve quickly - was correct and some arguments leading to this argument could have been correct, too.

When I was listening to the world's leading ID proponents in Nice in June 2010, I was sort of impressed by their knowledge and the seriousness of their proposals although, in my opinion, the "no-go theorems" always had loopholes because they imposed too many conditions on how the life was supposed to evolve (e.g. the mutation rate as a function of time), instead of allowing it to evolve in the near-optimized, constantly adapting way.

Dear Gene, apologies if it's because my not having spent enough time with your text but I don't understand your argument at all. Why are the two possibilities - miracle on Earth or hell of a lot of life out there - the only possibilities? Do you actually think that the previous text in your comment contains a justification of this bold thesis?

I think it's exactly the other way around. Life in outer space - where it evolved in the primordial forms - may be extremely rare. It may be 100 of spores per a Solar-like System but that's enough for igniting the evolution on Earth once the piece of dust with a spore gets to Earth.

Quite generally, I don't understand why you talk about planets. The point of panspermia was that the initial life forms evolved on some rocks or interplanetary or interstellar dust of a sort, not on planets.

Would the only way to know be to dig into asteroids around the Earth ? See this asteroid-mining project...

http://www.space.com/19368-asteroid-mining-deep-space-industries.html

Good luck but I don't think that panspermia predicts that life will be present in almost every asteroid. Quite on the contrary.

If X asteroids/comets/meteorites hit the Earth in 200 million years or so, X is a large number. But I think it's clear that panspermia only requires the number of cells on these X asteroids - a huge number - to be one or larger because it's sufficient to ignite life on Earth once it gets here.

So in my opinion, panspermia is doing perfectly fine if only 1 in a trillion of asteroids carries something like life! I would bet a lot of money that no non-Earth-related (not contamination from us) life will ever be found on asteroids.

Arguments (such as Ribo's) about the relative priority of prokaryotes versus eukaryotes are faced with the relatively recent (decades old) realization that eukaryotes are composed of components that look an awful lot like incorporated prokaryotes. The mitochondria that provide the energy for most eukaryotic cell functions are prokaryotes that still have part of their own DNA in them (the eukaryotic nucleus has just enough to exert some control over the rate of mitochondrial reproduction).

Choosing real dates for first prokaryotes and first eukaryote is tougher, since the "fossil record" is hard to read for such tiny soft organisms.

We really don’t know anything about how life begins and panspermia does not seem to address that question. Under favorable circumstances life might spontaneously arise very quickly or very slowly. It could happen in a time either short or long relative to a billion years but the fact that life came to exist on earth almost as soon as temperatures permitted the existence of liquid water suggests that life either happens quickly or that it is very rapidly transported to nearby places where continuing evolution is possible. In other words, it either pops up everywhere or spreads like mad. In either case it seems likely that there’s a lot of life out there. Of course the progression of life up to and including string theorists may be very slow, indeed.

This is all based on liquid water, which, I think, implies planetary origin, but is there any reason to believe otherwise? It also seems likely that protection from radiation is a necessary condition along with moderate temperatures. Of course periodic thawing might suffice.

The Wiki article talks about planetary origins and dispersion that begins with collisions. Am I misunderstanding panspermia? The liquid water limitation seems a bit arbitrary but I don’t know of any other life-friendly milieus.

Well, Lubos, the rarity or abundance of life remains an open question but I agree that the chances of finding life-bearing asteroids is remote

I'm not out to "disprove" panspermia here but I find the premises of the paper in question pretty spurious. To start off with, the authors suggestion that biological material would spread out by super nova explosions. Considering the violence of such an event, you'd not expect much apart from ionized gas to survive them. Also, the comets and asteroids you mention were formed together with the rest of our solar system out of micron-sized dust grains and molecular gas at extremely low temperature embedded in a harsh radiation field. Hardly a suitable environment for biological entities to survive. What we DO know is that rocks can travel inside our solar system when knocked off by a suitable collision, we find rocks from Mars on occasion. But the challenges of transporting a meteor across interstellar distances are pretty formidable. The impact that kicks them off would need to give them enough speed to break out of the stars gravity well and then you'd need the biological material to survive viable to replicate across megayears of high radiation, vacuum and harsh radiation. To do this while maintaining DNA repairing metabolism like the authors indicate would seem... Unlikely.

Then we have the idea of applying Mores law. This seems like total bogus. As an example; The species with the largest genome know is a plant that has about 50X the number of base pairs a human has. Either modern humans developed ~2Gy ago (which we didn't) and the plant in question popped up very recently (and is our superior) or the basis of the theory of steady exponential development is not true.

We also know that the number of genes only weakly correlate with "more complex". What seems to be more important is how efficiently a genome is used, call it information compression, if you like.

From what we know, genomes increase in size mainly by co-opting other genomes. Our own genes are stuffed with bits and pieces from retroviruses, as an example, which leads to:

Then there is the premise of some sort of steady, inevitable, development. From what we know, evolution happens more as bursts than as a steady process. We have large dyings that tend to lead to explosions of new species. This would be particularily important back when plants developed (which we know happened on Earth) and oxygen killed off most of the original batch of species that didn't manage to hide somewhere out of reach (wiping out those potential panspermie and "resetting" the slate). These rapid evolutions then tend to be followed by long periods when not that much is happening. This makes sense if you look at when there is a need to adopt (a crisis) and when it is ok to stay the same (stable conditions).

And of course, we just don't know what ancient genomes looked like. It is all well and good to look at procaryotes and eucaryotes (who btw have the same range of numbers of base pairs ;)) that live today. But what was the size of the genome of the first species that crawled out of the sea? We just don't know.

Whatever one thinks about panspermia, this quote from the featured paper is foolish hubris "there was no intelligent life in our universe prior to the origin of
Earth, thus Earth could not have been deliberately seeded with life by
intelligent aliens;" Talk about unsupported prior assumptions -- worse than Climate Science.

I do not have any problem with the panspermia idea, but I do have the problem with the graph, as genome size is not growing smoothly, bat rather in steps. For a long time there is bacteria, than they manage to recruit mitochondria and size jumps... Basically there are about 3 such a jumps in the history.

On the other aside in animals genome sizes range more than 3,300-fold, and in land plants they differ by a factor of about 1,000. Protists genomes have been reported to vary more than 300,000-fold in size. Bacterial genome size vary about 100 fold.

( look here for example http://en.wikipedia.org/wiki/File:Genome_Sizes.png)

Considering that, put the error bars on the line, and you'll see, you have no line. ;-)

Seems to me that you are saying life can be either rare to emerge but easily spread or just easy to emerge. As life on earth appears to go back 3.5B years, within at most a couple of hundred million years after the earth became cool enough for liquid water, either proposition could be true.

It seems to me the 'easy life' option is simpler, but short of sampling lots of other planets physically to verify related DNA structures, it is hard to see how panspermia could be experimentally verified or falsified.

More interesting imho is that 'intelligence' as we understand it seems to have evolved only once in those 3.5B years, despite repeated resets, aka mass extinctions. We may be alone in a universe teeming with life, based on our experience.

Wherever and whenever natural selection starts it presupposes an environment in which there is reproduction and competition as well as mutation. Can you imagine that happening in a gas (presumably enclosed in a solid, itself warmed by radiation or radioactivity) or on the surface of a solid similarly fueled? Also, assuming the process gets started with steps 1,2,3, etc. Do you imagine all the steps up to the beginning of life on earth occur in one place or would there be multiple independent panspermia seedings? Does the fact that the same small number of twenty or twenty-one amino acids out of a much larger set are used exclusively raise any obstacles? These are just thoughts off the top of my head so no doubt you've got answers. I think it is a neat idea if you do.

Here is a link on how water came to earth : http://news.nationalgeographic.com/news/2009/08/090805-earth-oceans-comets-life.html

Evidence for a comet swarm is given: "A barrage of comets may have delivered Earth's oceans around 3.85 billion years ago, a new study suggests."

Just in time for the graph, considering the errors.

I think this argument cannot go further until one can create life in the petri dish from the chemicals that make it up.

When DNA was discovered it fulfilled my expectation that everything that exists has a quantum mechanical probability and a DNA cell will form given the necessary ingredients and conditions, inevitably. We do not yet know the conditions to create live DNA from scratch.

So the panspermia hypothesis is viable as long as we do not know all the conditions, to decide whether it is a probable hypothesis or not.

The water comets though should help the model.

OK, I think that your criticism is based on the idea that these primordial life forms are very vulnerable and require a luxurious environment.

That's what I tried to claim is a basic, universal mistake in the thinking of people like you. I have clearly failed to convey this point.

Explosions of supernovae create some huge temperatures somewhere but they lead to tolerable temperatures further away and microbes like that have no trouble to withstand the accelerations and similar things. When one is this small, the Earth-like acceleration is totally negligible for its life.

You also complain that a micron-sized dust is a bad environment for this life. I disagree. It's arguably the best environment. For example, bacteria and fungi and similar cells have diameters of e.g. 3 microns. More primitive organisms were arguably even a bit smaller. So it's very convenient to live on pieces of dust that are comparable in size. Collisions with other similar pieces of dust may provide the living forms with the material they need and they may pick etc

Quite generally, you have an anthropomorphic imagination of the needs of the early life and I think it's manifestly wrong.

Dear Kip, it is not any independent assumption at all! It's just a consequence of the cosmic biological Moore's law they not only assume but also support with a great deal of evidence.

If one needs more than 3 billion years to create genome of the human proportions, then it follows that there couldn't be intelligent aliens 4.7 billion years ago who could have "seeded" anything.

The authors don't claim that they have any rigorous proof of their picture but they have surely provided much more evidence backing their claims than the evidence you provided to back some contradictory claims.

Dear Gene, I don't understand your criticism "we don't know how [quickly etc.] life begins". This is the very main purpose of this paper!

With some definitions of the "panspermia hypothesis", it may be silent about the very origin of life. Perhaps. But this universal genome Moore's law is not silent at all.

Life is evolving pretty much gradually towards larger genomes, the basic building blocks are the bases etc. which are molecules that may be found abiotically, and it takes a billion of years for the genome to increase by an order of magnitude.

What's wrong with that? What's missing in the explanation of the "origin of life"? One may study *specifics* of particular life forms much like one can study particular microprocessors from the late 1970s, and so on, but I wouldn't count it as "study of origin of life".

I thought the Arxiv blog was pretty interesting (I plan on checking out the paper soon.) If nothing else if gives you some fun ideas/numbers to play with and (at least for me) stirred up plenty of new and relation questions to mull over. One of the most interesting things in the blog post I thought you didn't mention: The genomic growth rate implies a genesis of ~10 billion years ago. This would actually have a bearing on the Fermi problem, because if it takes 10 billion years for sufficiently complex life to develop, then perhaps the (relatively) similar age of the universe (and exacerbated by early universe particulars) explains why the galaxy isn't (presumably) crawling with E.T.'s.

On a more critical note, though, doesn't the growth rate require continuous evolution? My point being that, e.g., 4 billion years spent as a frozen spore on an asteroid hurdling between stars wouldn't contribute at all, right?

That molecular building-blocks of RNA and DNA, nucleic acids, can and have formed elsewhere than on this planet is not a controversial idea - it is AFAIK a proven fact. This does not mean that the kind of environmental conditions for RNA-replication and later DNA-replication has a snowflake's chance in hell to get going in outer space! Almost everyone end up having some pet idea or belief; And so, IMHO, you, dear Lumo, could have one that is a lot more 'off the planet' than the innocent idea of panspermia! :-))

Here is a nicely relevant (IMO) webpage http://exploringorigins.org/nucleicacids.html

They use the measure of "non-redundant functional nucleotides" which should vary less.

Lubos, so you're saying that most earth-like planets should have life... which leads directly to Fermi's Parodox

A good point! But not necessarily.

What I'm saying leads to life on Earth-like planets in the stellar neighborhoods where the previous form of stellar life developed - which is not necessarily everywhere.

The hierarchies may exist at every level.

Unless you have a reason that earth occupies a privileged position, our "stellar neighborhood" should be common... and Moores-Law-Panspermia suggests that intelligent life should plentiful. I think the idea is cute, but it leads to the conclusion that some UFO sightings are actually robots from other systems... how do you reconcile that?

Couple of things here. As it relates to the Life before Earth paper and panspermia, it seems pretty trivial. They assume that genomic complexity will increase if an entity possesses several traits (gene cooperation etc...) but this is not surprising.

If i take a sterile environment with an appropriate food source and add a single bacteria and leave it for some time, the genomic complexity of that environment will increase exponentially. This fact does not tell us anything about how the bacteria came into being in the first place.

The point is, it is not very interesting to assume that an organism with some pretty complex features will behave analogously to Moore's Law. when you can't explain the existence of those complex features in the first place.

For myself, I cannot see any plausible origin of life on Earth in the absence of organic compounds and liquid water. I would assume that life on Earth originated on Earth because it has had the most liquid water on it for the longest time)in our neighborhood). It is certainly possible that life originated on Mars or whatever and migrated here but this does not seem too likely.

Interestingly enough, not only does the mutation rate vary in different organisms and within different portions of an organism's genome, but many organisms also have the ability to vary their mutation rate in response to environmental factors(see stress induced hypermutation). Organism's have different versions of DNA polymerase that are more or less effective at correcting DNA copying errors).

I don't think so. Even if we don't know what's special about the solar neighborhood and nearby stars, they may still be special for some reason just like the Earth is special when it comes to habitability.

Quite generally, I don't think that panspermia implies lots of ETs or robots - quite on the contrary, it's a framework in which these predictions may be avoided.

Dear Luke, I surely think that the competition for resources etc. is totally plausible on the surface of some micron-sized dust. After all, it looks like just a small copy of larger animals in larger environments. Gravity doesn't attract the microbes to the dust but the surface tension is probably enough to replace it.

It's a point of panspermia that one uses a large volume/surface of the outer space to try various alternatives in the early steps - but it's still plausible that the resulting life only arises in a very small fraction of the outer space and it's hard for it to propagate. Also, different layers of the early evolution may take place in different regions - or at least differently focused regions - as long as it is possible to transfer the organisms from one place to another.

There may be independent panspermia seedings, with different types of life etc. It should still be unlikely for two different sources to invade Earth or another planet.

I think that champions of this meme generally assume this process to occur within one galaxy simply because the speeds needed to transfer life to another galaxy seem prohibitive and linked to high temperatures that kill any life.

I agree with the point here that there's no reason why very primitive life would need a habitat so congenial as earth. Therefore it is not parsimonious to assume that early life evolved faster.

It seems too a wishful thinking Lubos, since actually just ~ 400 base pairs are needed to encode simplest life forms (http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/G/GenomeSizes.html) ... so supposing that life started on our Earth is still an excellent working hypothesis. Other one is that proteins developed first than RNA in the primordial sea and so catalysed the emergence of RNA: this explains why one doesn't see "computers with order of 1 transistors" ;-)

I am not sure what you are getting at but the fact is that "genome size" has no meaning in the absence of some kind of "living" context. Reality check here if we found a thousand ton asteroid that was a single double stranded DNA molecule, that would not imply anything about whether natural selection is taking place.

A system that is capable of undergoing natural selection is very complex, much more so than a genome on its own.

Furthermore, once NS starts, it will dominate over whatever pre-NS genome complexifying processes were taking place (see population genetics for evidence). As such, we have good evidence for what genomes look like when they are dominated by NS and virtually no evidence for when they are supposedly dominated by panspermia. Do we have any evidence at all of what the genome size was before the origin of the Earth?

In re: the water issue, it is true that there is lots of water in the Solar System but the vast majority of it is not relevant to the kinds of chemical reactions that take place in living organisms (those reactions need liquid water).

"the possibility that life could have evolved at many places before the Earth already weakens the need for a creator."

OTH it shows he didn't sit around for 4 or 5 billion years before he got started. :)

Dear Shawnet, the need for the living context is OK because it comes automatically with the molecule. You won't find a very long DNA molecule that is free of a living context!

DNA molecule not only determines what the living form looks like and does - but it's also a consequence of some production that needs proteins and other things, parts of the living form. Once you can do such things to produce DNA, you have a living context! Living forms able to do such things differ in functional complexity and efficiency of the storing but these are among criteria that affect the natural selection - so ultimately among the competing forms with the same length of DNA, the more clever or functionally complex ones, or those with a better efficient data storage in the DNA, will prevail! This is true on the Earth as well as outside the Earth.

LM

Lubos, I think you are underestimating the difficulty of getting a system that is capable of undergoing natural selection by a huge amount. DNA on its own is not a living context. DNA is a useful molecule precisely because it's molectules don't react differently when put into strings of a different order or size. As such, NS has nothing to feedback on. OTOH, the order of amino acids can cause big differences how the a string of them reacts with the outside world (and hence give a lot for NS to work with) but doesn't make copies of itself (except in pretty trivial cases), so there is no way for the changes made to last very long or be duplicated.

OTOH, if you can come up with a process that ties the storage properties of DNA(or equivalent) to the reactive properties of amino acids(or equivalent), then you are getting someplace. Unfortunately, as far as we can tell, this is very complex (no one has been able to figure out how this might have happened even in theory).

IAC, more to the point of the particular paper in question. It makes no sense to compare the growth of genomes before you get this sort of combined storage-reactive mechanism to the growth of them afterwards. Once this mechanism comes into existence, NS determines the properties of genomes but it can't do so before.

Cheers, :)

As I recall, R.A. Fisher in his General Theory of Natural Selection estimated that, on average, the chance of a favorable mutation going to fixity in a given population was about one out of fifty. In other words the same mutation at a particular point in a genetic string would have to recur about fifty times before it became the only allele at that location.

What this implies, it seems to me, is a rather large biota of genetic strings (primitive life forms) all multiplying like crazy and competing for space and available energy. Doesn't that imply an environment physically larger than a grain of sand? After all a lot of tiny worlds evolving independently don't sum. You need one connected environment of pretty good dimension to get much evolutionary "work" done. The question is, how big? Bigger than the surface of a baseball? A blimp? A small planetoid? It's a math problem, at bottom, a little like thermodynamics if I'm not mistaken, and I don't pretend to know the answer.

That's an interesting point. At least the "speed of evolution" scales not only with the total mutation rate, but also with some factor representing the competition for resources.

Dear Lubos, I like your blog. I just found it last night
when wondering what happens with the "lost energy" of the redshift
caused by the expanding universe. Then I had a great time looking at the other
topics you covered. Well done, and thanks. This time, as a biologist, I have
course but you neglected a crucial point: the fidelity of genome replication!
It is commonly believed, not by all, but certainly by most in this field that
early life was based on RNA. Swapping from an RNA to a DNA and RNA system would
be a relatively "easy" evolutionary step which may require only
little mutation in a RNA replication protein changing it's selectivity to DNA
building blocks. DNA is a more stable molecule and therefore a better place to
store information. Anyway, some RNA molecules, only about 160 base pairs long,
can catalyze their own duplication (some might even call this a very early life
form)! However, with a very bad fidelity of about 98% per base pair. So you can
see that the rate of mutations was very high. Later more complicated replication
systems based on RNA and proteins had a better but still not great fidelity. We
know this because RNA polymerase molecules never developed a proof reading
system (as is the case for DNA). This is why RNA-based genomes -like the flu
virus- can mutate so fast. Since evolution is directly dependent on the
mutation rate, it would have slowed down majorly after the
"prokaryotes", which is the first entry you have. To me it seems
that, after a DNA based proof reading system has developed, Moors law apply,
but not before.

Best wishes

Mich

Thanks for your kind words and interesting thoughts! Is there any evidence that RNA life took a long time at all?

My guess would be that exactly because the mutation rate is so high in RNA life, it takes a very short testing time in which it either dies or evolves into something more stable. So I would think that most of the time, the life on comets and dust etc. was a DNA life, too.

Also, I believe that the first DNA molecules that evolved after life switched from RNA life still had to be relatively short because the low fidelity of the RNA life doesn't allow one to make too long and complicated, useful yet reliable, RNA codes.

Terrence McKenna claims that mushroom spores can survive intergalactic travel and are likely the panspermia source