|
Author
|
Topic: Because it's hard - "more bang for your buck"
|
Micah Sparacio
Member
Member # 6
|
posted 09. May 2003 10:10
In the primacy of nucleic acids thread, brauer says the following:
quote:
In general I agree with you: there is an emphasis on DNA that perhaps discourages extensive investigation into other biochemical units of organization.
But this fact illustrates quite nicely the concept of methodological versus ontological assumptions. We work with DNA because right now it's easiest. We don't work with other cellular structures (membranes for example) that may contribute to inheritance because it's hard. It's a "more bang for the buck" kind of thing.
Most molecular biologists don't think about this much: they simply work in the realms in which they think they have a good handle. If questioned seriously, they'd admit that DNA is not necessarily the whole story. But from the perspective of a practitioner, it doesn't really matter: our technology makes it convenient to make the assumption that it is for methodological reasons.
I think that this is a good point. We should certainly put our research work where it will pay off the most. Not being a practicing biologist (and certainly not understanding the nuances of the issue), however, a question naturally arises. If it is hard to investigate "other cellular structures (membranes for example) that may contribute to inheritance" and we put our resources into genetics out of utility, do we run the risk of developing our technology and theories around genetics to the detriment of making it easier to look at these other aspects of biological inheritance?
Putting this in computer science terms, if we consider science as a search algorithm (which in many ways it is), do we run the risk of getting stuck in a local maxima?
For the computer scientists out there, what search methodology do you think the scientific community currently uses and is there a better one to consider?
[ 09. May 2003, 10:12: Message edited by: Micah Sparacio ]
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 09. May 2003 13:23
Micah wrote quote:
Putting this in computer science terms, if we consider science as a search algorithm (which in many ways it is), do we run the risk of getting stuck in a local maxima?
For the computer scientists out there, what search methodology do you think the scientific community currently uses and is there a better one to consider?
I think this is a fairly good analogy. To expand it a bit, what I read brauer to have said is that the current modal tendency is to work on DNA-related research - there's a clustering around that 'maximum.' Fifty years ago that wasn't the case. After Crick and Watson discovered DNA's structure, and especially within the last 20 or so years when the various technologies for analysis became faster, cheaper, and better understood, there was a 'flow' toward DNA research in the population of researchers. The availability of research technologies and instrumentation are real important drivers of that modal tendency.
But there's also considerable diversity in the population of researchers, with smaller clusters distributed around the 'research topic' landscape, each cluster working in other areas of emphasis. As, if, and when they find and/or invent research technologies, and as, if, and when the reward structure (fitness function) deforms as it inevitably does with new knowledge, then one will see 'migration' on the landscape. Actually, one will see what appear to be near-saltational jumps - cultural evolution is not as constrained as biological evolution, so the metaphor begins to break down badly in describing the system's dynamics.
While diversity exists, the system is not likely to be (permanently) trapped on a local maximum.
RBH
IP: Logged
|
|
Art
Member
Member # 179
|
posted 09. May 2003 22:33
Micah said: quote:
I think that this is a good point. We should certainly put our research work where it will pay off the most. Not being a practicing biologist (and certainly not understanding the nuances of the issue), however, a question naturally arises. If it is hard to investigate "other cellular structures (membranes for example) that may contribute to inheritance" and we put our resources into genetics out of utility, do we run the risk of developing our technology and theories around genetics to the detriment of making it easier to look at these other aspects of biological inheritance?
There's a different way to look at the current trend to DNA-centered research, and it may tell us something more fundamental than simply "it's easy".
Certainly, in this day and age, it is feasible to do lots of research that does not involve DNA. However, "DNA-based" methods expand the scopes of our efforts, rather than limit them. For example, if many of us were given a problem, say, involving membranes, the most logical (not necessarily the easiest, but probably the most informative) approach would be to alter the membrane in as close to a physiologically-relevant state as possible. This, in turn, takes us to "the DNA" - either via standard genetics approaches, recombinant DNA and transgenics, reverse genetics, and the like (just to name a few jargon-laden methods - sorry about that ![[Smile]](smile.gif) ). This is, I believe, the true lure, the real power, of "DNA-based" strategies. It's not that it's necessarily easier, but rather that the answers bring us closer to truly relevant models and understanding. (I would add that I am alluding to broad-based studies, in which genetic and molecular approaches are complemented by more standard biochemical, cellular, and/or physiological experiments.)
All of which raises an interesting counter to the nascent idea that there is something more than "the DNA". For if "DNA-based" strategies can in fact be so useful in studying phenomena that are, at face value, clearly distinct from DNA, it would seem to suggest that "the DNA" probably ties in more intricately to most aspects of the organism that is immediately apparent. IOW, if we can use "DNA-based" approaches (sorry for all the quote marks, but I am too lazy to replace the euphemism with a more accurate term) to study virtually all manner of biological phenomena, might it not be likely that "the DNA" actually is central to all of these things?
Just another side to the discussion. Or, if nothing else, a plug for the power of multifaceted model systems.
IP: Logged
|
|
Mesk
Member
Member # 630
|
posted 12. May 2003 03:52
While molecular biology (A.K.A. "DNA-based research") is certainly the most active field at the moment, other biological macromolecules are hardly falling by the wayside. Indeed, proteins are becoming more and more of a focus (the so-called "proteomic revolution"), especially with advances in the techniques used to study them. New experimental approaches based on old technologies, e.g. mass spectrometry, liquid chromatography, two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) and the yeast two-hybrid system are already doing for protein studies what PCR and microarrays did for molecular biology, and I can guarantee that even more revolutionary techniques are just around the corner.
Many young scientists like myself are hedging our bets by learning whatever we can about all the techniques available, be they DNA-based, protein-based, or simply computer-based. Broad technical knowledge will (IMO) become ever more important as increasing numbers of scientists begin to specialise in one field or another, because those who are able to coordinate efforts in more than one area will become more and more useful. A multi-disciplinary approach is invaluable for modern biological research - for instance, my own research project draws on epidemiology (genetic association studies), molecular biology (gene cloning and recombinant DNA technology), protein biochemistry, animal genetic engineering, cell and tissue culture models, and a vast array of computer-based bioinformatic techniques. This allows me to follow a biological "story" all the way from genotype to phenotype, with plenty of stops along the way. When all of the techniques can be well coordinated this is an exceptionally powerful approach.
With respect to other areas of cell biology, such as lipids, I suspect that research will only really start to explode when new techniques become available. Membrane research is notoriously difficult, and is regarded by most biologists I know as pretty tedious stuff compared to, say, proteomics - largely because lipid research is so difficult that progress is comparatively slow. I am sure that there is a vast amount of potentially paradigm-shattering information about cell membranes that we are yet to discover, but figuring out what this information might be will require some serious technological advances (and, of course, some serious funding).
Mesk.
IP: Logged
|
|
|
Printer-friendly view of this topic