A Case of “Maybe-Science” — Questions regarding a Computed “Fossil Genome”

22 December 2010

I invite a rebuttal to what follows — it may be that I’m being unfair and an idiot at the same time

You can use the “Contact” button on the left side of this page to initiate the conversation. I’m always eager to learn.

Is no one else suitably skeptical of this “computed-fossil genome”?

Two scientists claim to have come up with a computed “genomic fossil” that allegedly recreates something that might have lived 3 billion years ago.

The work is tactically brilliant, but its strategic assumptions are obviously flawed.

Somewhat harshly put (to make the point) — these researchers substituted (i) an ignorance-based model of genetic evolution (ii) for the almost completely unknown evolutionary reality it supposedly is investigating and (iii) then inferred that their result says something accurate about something other than the workings of the model itself.

A brilliant conception, with a probably flawed result

RedOrbit.com (a science news website) said this about Lawrence David and Eric Alm’s Nature article:

Scientists at the Massachusetts Institute of Technology (MIT) built a “genomic fossil” -- a mathematical model that took 1,000 key genes that exist today -- and calculated how they evolved from the very distant past.

RedOrbit staff, 3 Billion-year-old Fossil Described, RedOrbit (20 December 2010)

From the scientists’ abstract:

Here we use an explicit model of macroevolution including gene birth, transfer, duplication and loss events to map the evolutionary history of 3,983 gene families across the three domains of life onto a geological timeline.

Surprisingly, we find that a brief period of genetic innovation during the Archaean eon, which coincides with a rapid diversification of bacterial lineages, gave rise to 27% of major modern gene families.

A functional analysis of genes born during this Archaean expansion reveals that they are likely to be involved in electron-transport and respiratory pathways.

Genes arising after this expansion show increasing use of molecular oxygen (P = 3.4 × 10⁻⁸) and redox-sensitive transition metals and compounds, which is consistent with an increasingly oxygenating biosphere.

Sounds pretty impressive. Especially given the uncertainty-concealing certitude of the article title.

(The misleading title is one reason that I don’t feel like a completely nasty-person in questioning the authors.)

What David and Alm implied

Not only are David and Alm implying that they know:

(i) which genetic life form spawned

(ii) the many species that followed

and

(iii) what proportion of “his” genes live on in them —

(iv) (this despite the fact that science has recently discovered just how few species human beings have actually identified,

and

(v) how many of the undiscovered ones live in unseen or harsh environments presumably more similar in functional requirements to model-critter’s, therefore making them still more relevant, yet unavailable, to the model) —

but also

(vi) when “model-critter” lived,

and even

(vii) very roughly how and when model-critter’s progeny affected electron transport and the atmosphere.

Another lay citation — with a revealing diagram

Wynne Parry, Genomic Fossils Reveal the Expansion of Ancient Life, LiveScience (21 December 2010)

David and Alm’s findings diagram is brilliantly done, but misleading

Lawrence David’s diagram (attached to Wynne Parry’s LiveScience article) summarizes the pair’s findings admirably well.

Its only pedagogic flaw is the visual imputation of certainty to something that cannot possibly be so clean.

The critical part (for the literal minded) — there is no actual fossil

There is no actual fossil.

Which ScienceDaily, for example, apparently misunderstood or semantically confused with its title, Three Billion-Year-Old Genomic Fossils Deciphered (20 December 2010).

Consequently, there is no actual genomic starting point 3 billion years back. The absence of fossils from so long ago is what (appropriately) inspired the David-Alm model.

This lack of a genetic starting point means that David and Alm:

(i) used a mathematical model of evolution,

(ii) based on living organisms,

(iii) that supposedly recreated a genetic-source “fossil,”

(iv) never once actually seen or “collected,”

(vi) it might have existed approximately 3 billion years ago —

(vii) by using existing mechanisms and change rates

(viii) to infer mechanisms, change rates, and compositional structures across unimaginable periods of ever-changing geologic/biologic time.

In short, the model determines its originating material based solely on its own assumptions

This technique casts today’s (poorly understood) mechanisms and “genetic-packets” backwards to create a starting point that has to be almost completely dependent upon the assumptions incorporated into the model.

In other words, the model and its result are almost certifiably off-the-mark. Or they are “nonsense,” if you’re impatient with the idea of piling multiple layers of assumptions atop one another and calling the method’s result science.

There is currently no way to be even vaguely certain that the David-Alm model’s (a) genetic mechanisms and (b) the rates of change attributed to them, remained consistent (or even remotely quantifiable) across that span of time.

Almost certainly, they did not. Genetic change does not “tick” predictably like radioactive elements do.

Creating such a hypothetical fossil might make generalized sense, until your recognize the largesse of our ignorance

On its face, the David-Alm approach might seem reasonable.

But then consider that we still don’t know how DNA, RNA, and epigenetics actually do what they do. Science recently recognized that we don’t even understand the minutiae of the division of labor between them. Indeed, only recently have the vital roles of epigenetics and non-coding RNA gained attention.

We have also discovered that some genes, earlier thought to have been conserved and ancient (due to their importance in function), are of surprisingly recent origin.

Researchers are also confused about the role(s) that allegedly “silent” genetic material plays. We increasingly suspect that at least some of this DNA is not really silent. And recent discoveries indicate that much of the material that may not be silent seems (nevertheless) to have been rather randomly generated.

In short, we human beings are Ignorami (with a capital “I’) about how genetically active material does what it does and in what proportion.

We can’t even say, with any degree of certainty, what genetic features correlate to which expressed manifestations.

Most importantly, we are not even sure what is genetically active material and what is not.

Atop all this ignorance, lies the vacuum of knowledge as to how organisms (i) we’ve never seen, (ii) affected each other’s genetic material and evolution, (iii) during 3 billion years of changing environments and their ecological “critter-sets.”

Hmmm . . . .

In sum, are we to accept (with a straight face) that David and Alm:

(i) turned this egregious pile of contemporary ignorance

(ii) into a “model”

(iii) that accurately operated across more than 3 billion years

(iv) of dramatic planet and biospheric change?

I doubt it.

But, in fairness to Drs. David and Alm, I’ve over-stated the doubts

The study author’s sound defense is that they tracked existing bits of genetic material backward. It is irrelevant whether we completely understand (i) what the bits do or (ii) whether they needed other (currently unknown) bits to function.

The fact that the study’s genetic material has provably been conserved through most of post-Cambrian evolution reasonably indicates that science can use them as markers for change, even before the trackable record begins.

On the other hand, maybe I haven’t over-stated

One of the implications I’ve drawn from medical science — which often concerns roughly the same kind of genetics that David and Alm are concerned with, particularly in microbial research — is that correlations between “bits” (of allegedly genetic material) and their effects are often completely wrong.

The marker is usually not the territory. And the marker-bit is (frequently enough) ultimately shown to have nothing at all to do with even tagging the territory or phenomenon that we are attempting to track.

If that’s true, then my doubts are probably not overstated. At least not by an order of magnitude.

At base, I am profoundly suspicious of too-easy solutions to the usually overwhelming complexity of biological phenomena.

Casting changing complexity backward through 3 billion years, and claiming to come up with a potentially reasonable result, seems too big a leap to me — at this stage of human knowledge.

Biology, evolution, and genetics are not physics (or physics’ related disciplines)

Biology and evolution are not the equivalent of physics or its offshoots.

Physics (via cosmology/cosmogony) can almost literally look backward in time, given the nature of light and radiation.

Biology does not yet incorporate moderately well-supported theorems equivalent to the quasi-rules of physics. The David-Alm model cannot compare to the defensibility of statements about the history of the recognized universe.

Consequently, I’ll take the David-Alm findings with a whole box of salt.

A useful rule from a brilliant Physics professor

Dr. Ernie Rost, my favorite Physics professor (now deceased), once told us (paraphrased):

“When you come up with an answer, take a step back to think about whether its units and magnitude make sense.”

Rephrased to fit here:

“Consider whether your factual base and your theoretical underpinnings justify the perceived specificity of the answer you propose.”

PeteFree.com

A Case of “Maybe-Science” — Questions regarding a Computed “Fossil Genome”