Biocurious

/ a biophysics blog

A Unified Theory of Biology?

Posted 2 September 2006 by Andre under &

Mohammed AlQuraishi, a graduate student studying genetics at Stanford, just wrote a guest post at Lubos Motl’s Reference Frame “On a Theory of Biology”. He discusses, but doesn’t entirely distinguish, two related ideas: the role of theory in biology and the possibility of a Theory of Biology. I would argue that the first is an important and vibrant field while the second either has existed for a long time or is impossible, depending on your perspective.

Regarding theory in biology, I think it’s important to emphasize that there has already been a variety of significant work: single molecule mechanics, protein folding, membrane biophysics, cell mechanics, cell motility, regulatory networks, and neuronal networks just to name a few! When the topic of theoretical biology is raised it is frequently thought of as something that will happen in the future, once we have that elusive “next generation” of quantitatively trained biologists, but it is happening now and those new to the field (like me) ignore that fact at their own peril.

So what about a Theory of Biology with a capital T? As is often the case in these discussions, physics is held up as a model to be emulated and it’s certainly true that there is a close interplay between theory and experiment in physics, but is it really as unified as is often claimed? I’m not thinking now about quantum gravity, where unification seems like a matter of time (!), but about condensed matter where the situation is more complicated. In one sense, low energy physics, including all of condensed matter, has been unified since Schroedinger, but as Bob Laughlin likes to make painfully clear to his graduate students [pdf], you can’t derive superfluidity from Schroedinger’s equation. Although consistent with quantum mechanics, huge areas of condensed matter, especially soft condensed matter, have essentially nothing to do with it. The situation is similar in biology but even more extreme. Schroedinger’s equation is also, some ideas about homochirality and consciousness notwithstanding, a unified theory of biology, but what is it good for? Well, small scales. But it doesn’t really provide much insight into the larger scale workings even of a bacterium.

Which brings me to my last point. A single meaningful Theory of Biology is not possible. I don’t discount the possibility of more powerful theories of many aspects of biology such as increasingly broad and quantitative “laws of evolution”, or the eventual quantitative understanding of a cell, but a single theory, expressed as a usefully small number of equations or a usefully short algorithm, that accounts for biology from the level of molecules to populations will never be. This isn’t a bad thing, the same way it’s not a bad thing that some equations can’t be solved explicitly. We just need to find other ways that we find satisfying to account for what we observe. Also, this doesn’t mean we won’t have a series of overlapping effective theories that apply at different scales that will ultimately cover the territory from molecules to populations. In fact, I think we’re well on our way down that road.

Instead of a single theory, what we should keep our eyes open for, although not exactly look for—science doesn’t seem to work that way—are new principles that cut across boundaries and can inform theories at every scale like evolution does. Some equally profound principle would be hugely exciting and might still be discoverable, but it won’t be a Theory of Biology.

  

  1. Doug    3 September 06    #

    Perhaps one will need a Category of Biology Information, composed of many theories.

    Since evolutionary biology is basically a multiple series of continuous transformations with radial and spiral symmetry that changes with either reversal, folding or breaking, such a system might be represented by some type of Lie Algebra and Lie Group.


  2. Xerxes    5 September 06    #

    This argument against reductionism always strikes me as very silly. Presumably the speaker intends to emphasize that

    You CANNOT derive superfluidity from quantum mechanics.

    But really, it is merely the case that

    YOU cannot derive superfluidity from quantum mechanics.

    Surely, (under the assumption that quantum mechanics is a correct underlying theory) it must be possible for an arbitrarily powerful computer to derive superfluidity from quantum mechanics. The inability of our current computers or analytical techniques to do so seems to me to simply point out the inadequacy of humans rather than of quantum mechanics. After all, superfluidity could not exist if Nature could not derive it from quantum mechanics all the time.

    Anyway, just because you understand a fundamental theory doesn’t mean you can get anything useful out of it at the scale you want with the computing resources you have. You just have to come up with effective theories that work where you want them to work and stop caring whether they work in places where you’re not very interested in the results. Even if we do finally reach the most fundamental theory of nature, the derivation of new effective theories in fields like biology will keep humanity and its expanding computational capability busy for an indefinitely long time.

    So, to get back to the original question, the underlying theories of biology (quantum chemistry) is already pretty well known, as are the further underlying theories (atomic theory, QED, the Standard Model) for several levels. But you wouldn’t call this a Theory of Biology, since it’s useless as an effective theory. The notion that an effective theory exists that is simultaneously useful across all the varied subfields of biology seems very silly.


  3. Andre    5 September 06    #

    Hi Xerxes,

    From the tone of your comment I get the feeling you meant to disagree, but I don’t see what your getting at. In fact, your last paragraph could almost be a summary of my main point.

    About superfluidity, I didn’t mean that it’s impossible in principle to simulate, although I see I might have given that impression. Just remember the importance of your use of “arbitrarily powerful” though. Here’s another quote from Bob Laughlin
    “Consider, for example, the celebrated spin glass problem – a set of N half-integral quantum spins interacting by random Heisenberg exchanges. Because this system’s configuration space has dimension 2N, a straight solution of the quantum mechanics requires the diagonalization of a 2N x 2N matrix, something we know how to do algorithmically. However even for the case of N = 200 this matrix has 2400 = 2.6×10120 elements, a number vastly larger than all the atoms in the visible universe. The computational task would obviously exceed the memory capacity of any conventional digital computer that could ever be built and is therefore fundamentally impossible [...for a classical computer].”


  4. Andre    5 September 06    #

    2400 = 2^400 and 10120 = 10^120


  5. Xerxes    6 September 06    #

    Well, sure. The quote points out its own flaws: the prediction is impossible using “any conventional digital computer that could ever be built”, which he says is “fundamentally impossible”. But I say it’s just physically impossible, not logically impossible. Thus, the explanatory power of the fundamental theory is not diminished. It is limited by our available resources to interpret it, not by its inability to explain particular phenomena.

    I guess it depends on what you mean by “explain”. To me, fundamental theories do actually explain all larger-scale phenomena; that we cannot personally construct computers that give us predictions we like is hardly relevant to the issue of explanation. On the other hand, we would rather have a theory that is slightly wrong and produces useful results than one that is exactly right and impossible to use. So if “produces useful results” is what you mean by “explain”, then fundamental theories don’t do that.

    So I do mean to agree with your general tone and point (that there is no and never will be a useful Theory of Biology), but I mean to disagree with the specific point that condensed matter phenomena are not explained by quantum mechanics.


  6. PhilipJ    6 September 06    #

    You can derive superfluidity from first principles?

    The N spins problem described above is, I agree with you, a physical impossibility, but not a logical one. But unless I’ve missed some extremely exciting results in condensed matter physics, the superfluid argument remains.


Name
Email
http://
Message
  Textile Help