|
ONLY CONNECT
In a report this week (Nature, Vol. 406, Issue 6792, July 13, 2000) that could help set the agenda for the post-genomic era, George von Dassow and colleagues from the University of Washington, Seattle, use computer modelling to explore how interactions between genes form defined structures in living things. They focus on the network of genetic interactions that defines body segments in insects such as the fruit fly Drosophila. This system, involving genes such as wingless, engrailed, hedgehog and patched, is relatively well known. But even a simplified model of it required 136 equations with more than 50 free parameters for such values as the half-lives of molecules, their rates of diffusion between cells and the strengths with which they bind to the targets - almost all of which are unknown. The researchers then tried to determine which set of parameter values would produce the kind of segmentation that we see in real life-as opposed to other patterns or a complete mess. They expected that the system would be fragile, and would only work with a very limited range of parameter values. In fact the system did not work at all. So the researchers introduced two 'fudge-factors', representing molecular interactions only poorly supported by experiment. The result was dramatic. The network generated fly-like segments, and did so for an wide range of parameter values and input conditions. This may explain why insects that develop quite differently from flies (locusts, for example), and even completely different animals such as vertebrates, have developmental 'modules' that resemble the segment-polarity network of Drosophila. This kind of modelling approach also has predictive value, helping researchers come to grips with the flood genomic data now upon us. Peter K. Dearden and Michael E. Akam of the University of Cambridge, UK, discuss the work in an accompanying News and Views article. CONTACT: George von Dassow tel +1 206 543 9147, fax +1 206 543 3041,
e-mail dassow@u.washington.edu Michael E. Akam tel +44 1223 336612 , fax +44 1223 336679,
e-mail akam@mole.bio.cam.ac.uk (C) Nature press release.
Message posted by: Trevor M. D'Souza
|