The domestic dog genome

Within some of these buildings it seems that time has taken a giant step backwards. The figures bent over their computer consoles are engaged in much the same quest as the frock-coated microscopists and Dodo-stuffers of bygone years; they are today's taxonomists, though their raw materials are of course sequences rather than morphological features, and statistical models and ingenious computer algorithms have replaced fixatives, stains and intuitive systematics. Taxonomy is a noble subject, but it is the bedrock from which conceptual leaps can be made and not in itself what the Serbelloni participants would have called theory. Perhaps they would have felt their goals were closer to those of that other constellation of theoreticians, whose housing arrangements are also undergoing an expansion: the systems biologists. These brave spirits seek, amongst other things, to bring numerical prediction and dynamic modelling to cellular machinery. If they succeed, we should be able to interrogate a computer program and discover how a living system will respond to any conditions we give it. The system in question might be a special pathway; in moments of hubris, it could be a whole Escherichia coli cell. I have sampled my colleagues' opinions and found a solid consensus of opposition to systems biology. Mathematical models of complex biochemical systems have never worked, they say. Far better to go for logical, diagrammatic, piece-by-piece descriptions; then we will have the large picture. The more eminent the colleague, the more confidently he or she asserts this opinion. So many wise heads must surely be wrong. Unfortunately, the arguments on their side are rather good. Take the transcription factor network of E. coli. There are about 300 of these factors and an extrapolation from the hundred or so currently studied suggests the network will be very richly interconnected, with much feedback and multiple control of genes. This, however, is the least of the problem, for the operation of these factors can also be stochastic (there may be only a few copies of any factor in the cell), can depend on the formation of elaborate complexes, and can have a baroque mode of control of transcription, involving twisting or looping out of DNA, for instance. So what seems at first sight like a mere matter of several hundred rate equations turns into labyrinthine nightmare. That's biology for you. Pessimistic as this picture is, it is not as bad as the alternative. Listen to …