🚀 go-pugleaf

> >Q.What other ways could they test their products?
> >A. 1. Enough is known about toxicity in humans to put the information into
> >      computers and let the computers tell us if a substance or mixture 
> >      would be toxic.
> Sci.med: Is knowledge of toxins and biochemistry really that advanced?
> Is this just another expression of the myth "Computers can do Anything
> when given a lot of Data"?

	The truth is that if anyone ever tells you that they can absolutely
predict the function of its molecule from its structure, laugh in his
face.  
	Take the following two examples.  By a structure-function
relationship, Thalidomide should be perfectly safe.  As a matter of fact,
it is perfectly safe - in rabbits, and in humans beyond 20 weeks gestation.
No one ever suspected that it would be such a potent teratogen for a
short period in embryogenesis.  The FDA never approved it for use in the
US, but the tragedy in Europe was of immense proportions. It inhibited
the growth of the proximal limb bud, but who could have ever predicted?
	Or take Dioxin, this should also be a fairly innocuous compound.
It does not induce mutagenesis in bacteria and is fairly stable.  Rats
and certain strains of mice can literally eat it by the pound, but just
a few millionths of a gram will kill a guinea pig (or is it a gerbil).

	This is not to say that certain animal testing isn't excessive,
but the shear fact that biology is stranger than nature will ensure
that it will never be possible to eliminate it completely.

-- 
			      Craig Werner (MD/PhD '91)
				!philabs!aecom!werner
              (1935-14E Eastchester Rd., Bronx NY 10461, 212-931-2517)
                     "Well that's my story, not that it matters..."