a little extrapolation is a dangerous thing

The other day one of my friends sent me a link to this discussion of a recently published paper. (‘Published’ in the sense that it’s available through archiv, which I gather means it hasn’t been through peer review.) The actual paper is available here. Basically, the authors claim that life has increased in complexity – they’ve used genome size as their measure – as it’s evolved, and that extrapolating that trend backwards suggests that life evolved prior to the formation of the solar system.

But is genome size a particularly good proxy for complexity? Here’s the graph that underpins the conclusions reached by Sharov & Gordon:complexity vs time.jpg

Do you see what they’ve done there? ‘Worms’ – which worms? For after all, there are a lot of them: at least 10,000 species of flatworms, more than 80,000 species of roundworms (aka nematodes), and another 10,000 or so annelids (including the familiar earthworm), not to mention the less familiar taxa such as velvet worms & the priapulids. As for the arthropods – well, good old Daphnia has more functional genes than we do. (The poetical Cuttlefish has a nice take on this story here.)

And I see that plants & protists have been left out altogether – unless they’ve been lumped in under the general heading ‘eukaryotes’. Which is strange, because the overall genome size varies by 5 orders of magnitude** across the eukaryotes so far studied, so using a whole bunch of data points instead of the collective average, would make more sense. Unless that would spoil the nice straight line? (**Having said that, much of that variation is due to the number of introns & the quantity of non-coding DNA; however, the various regulatory sequence regions must surely come under the authors’ heading of ‘functional non-redundant genome’?)

I had also thought, on reading the review, that we were probably looking at an argument for panspermia. And I was right. This and other conclusions are presented in the abstract, & I note a certain amount of hubris in the assumption that humanity represents the only possibility of intelligent life in our universe (my emphasis).

(1) life took a long time (ca. 5 billion years) to reach the complexity of bacteria; (2) the environments in which life originated and evolved to the prokaryote stage may have been quite different from those envisaged on Earth; (3) 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; (4) Earth was seeded by panspermia; (5) experimental replication of the origin of life from scratch may have to emulate many cumulative rare events; and (6) the Drake equation for guesstimating the number of civilizations in the universe is likely wrong, as intelligent life has just begun appearing in our universe.

A.A.Sharov & R.Gordon (2013) Life Before Earth arXiv:1304.3381v1  

PS Strangely, in a paper supposedly about biological evolution, the latter part of the article goes on to discuss technological (ie cultural) evolutionary change – I’m not convinced that it’s appropriate to segue between a claimed link for genetic complexity & time, into the undoubted complexity and rapid ‘evolution’ of technology; apples & oranges, guys.

8 thoughts on “a little extrapolation is a dangerous thing”

  • herr doktor bimler says:

    So worms, fish and mammals are *not* Eukaryotes? Looking up ‘yeast’, I see that the S. cerevisiae genome has about 12 million base pairs, which doesn’t fit neatly on their graph.
    ‘Worms’ – which worms?
    Nematodamorphs? Priapulid worms? Sipunculids? Nemerteans?
    Saying that “worms” originated 1 billion years ago is just silly. So did non-worms!
    The implicit argument here is that as soon as chordates emerged from the other phyla around, back in the Pre-Cambrian, all those invertebrate phyla stopped evolving and speciating so their present-day genome is a snapshot of the genome size back then. Then tetrapods emerged from fishes, and all the other fish genera stopped evolving and speciating, and *their* genomes went into a time capsule.
    Nice way to beg the question, guys!
    So it’s the anthropocentric ladder-of-evolution fallacy compounded with the fallacy of “evolution stops for everything that’s not on our line of descent”.

  • herr doktor bimler says:

    “Prokaryotes” as a single point on the chart?
    Again, the authors seem to be assuming that once parokaryotes came together as working entities, things stopped happening to them (other than the ones that invented endosymbiosis and a nuclear membrane and became eukaryotes)… there was no further diversification, they haven’t changed or optimised their DNA in the subsequent 3-1/2 billion years
    “The size of Bacterial chromosomes ranges from 0.6 Mbp to over 10 Mbp, and the size of Archael chromosomes range from 0.5 Mbp to 5.8 Mbp.”
    Big question, too: how do they justify *wanting* to extrapolate back before prokaryotes? There seems to be a quasi-ID notion here that prokaryotes must have been preceded by simpler organisms with only 100 000 base pairs, and before that by 10 000 base-pair organisms, and so on, each stage taking the same time to evolve from its predecessor, but that makes no sense so Godddidit.

  • Alison Campbell says:

    I think greater minds than mine (ie PZ’s commenters) had also noticed the possible ID-thinking inherent in the paper.

  • herr doktor bimler says:

    I haven’t looked at the Pharyngula treatment.
    Biggest problem, perhaps, is that by the authors’ own logic and data the Complexity / Time graph is not smooth. It’s a series of discontinuous steps, which they have left undrawn, leaving the reader to interpolate between the points and imagine intermediate stages between (say) prokaryotes and eukaryotes. Or between unicellular and multicellular organisms.
    3 or 4 billion years ago the first prokaryotes appeared… and then for a long time not much happened. Then eukaryotes came along… and happily eukaryoted quietly among themselves, for a long long time, until someone invented the membrane proteins for inter-cellular signalling, and suddenly it made sense to be multicellular.
    So it’s a bit cheeky for the authors to compare the idea of a sudden step at the start of life — when diffuse chemical hypercycles managed to encapsulate themselves as cells — to the abrupt jumps of Creationism.

  • Alison Campbell says:

    Yes, well, these many issues are presumably why the piece appears in arXiv, rather than a mainstream genetics journal…

  • Hi you two 🙂
    I never did find time to read the paper itself (perhaps a good thing?), but that graph looked, to be polite about it, questionable from so many angles. My original comment over at the MIT Tech Review site pointed at the curiously “linear” thinking about evolution too, but for some reason I canned that bit when I posted my comment. I agree it looks curious. The extrapolating back has to stop at cellular life with protein-coding genes, even if you overlook the “punctuated” aspects, so it can’t be used to go right back to the origins of life. I had a quick look at PZ’s place but didn’t see a piece on this – haven’t time to look too far back, so I’m guessing it’s older.
    re “are presumably why the piece appears in arXiv, rather than a mainstream genetics journal”
    Some people are encouraging biologists to use arXiv in the same ways as physics does, pre-publication peer-review and all. It’ll be interesting to see how that fares. I know a few people in bioinformatics have done this, but conceptually that field is closer to maths/computer science who use arVix regularly.

  • Any science student quickly learns that to make results look better, they should
    1. Leave off any indication of uncertainty
    2. Plot them on a log graph
    3. Stick a straight line through them even when one is not appropriate.

Leave a Reply

Your email address will not be published. Required fields are marked *