In article <khurley.803704818 at quapaw.astate.edu> Kevin E. Hurley,
khurley at quapaw.astate.edu writes:
>the risk of provoking response due to controversy, I am under the
>that "wet" experiments may yield sensitive/accurate data while software
>simulation will produce precise data exceeding that of any "wet"
>experiment. This precise definition yielded from simulation will push our
>theory, and is therefore useful and necessary but maybe not sufficient.
IMHO, a model simulation is an extremely useful tool, but with an
important caveat: the structure of a model is 100% defined, whereas more
often than not, we don't know exactly how the aspect of biology we are
modeling is defined. Models are almost always very simple approximations
of biology (and a lot of the time, not even very good ones). If we knew
100% how anything in biology worked, then we could stop doing "wet"
experiments on it, and just play around with the model instead. I don't
think this is the case, though. Probably, the most useful thing about
modeling is that when a model *doesn't* reproduce a particular "wet"
result, it tells us where there's a hole in our understanding that we
need to fix by doing more experiments and refining the model.