CDR: reverse engineering pedigrees (was Re: why should it be

[David Honig]

At 12:20 PM 10/19/00 -0400, Trei, Peter wrote:
>Assuming, of course, that the birth records accurately reflect parentage. 
>If you take a course in human genetics you're likely to be astonished at the
>
>rate of fooling around that must occur to account for the appearence of
>traits
>within families - I've heard that as high as 10% of firstborns must have had
>a father different than the one on the birth certificate (no, I can't give
>you a
>cite).

When I was reading about other nations' plans to do what Ken Brown pointed
out only Iceland does now, it occurred to me that even with 
an anonymized database, you could infer parentage from the raw
anonymized data -its a harder problem than testing a particular
parenting hypothesis (is A likely descended from B + C, given
samples of A,B,C?), but maybe possible.  

Maybe the EULA will have a no-reverse-engineering-pedigrees clause..