URGENTLY Seeking Statistical Reference

wijsman at max.u.washington.edu wijsman at max.u.washington.edu
Sat May 8 02:33:29 EST 1993

(stuff deleted)

> The study involved taking blood samples from mothers, fathers, and
> children, and conducting some sort of analysis (what kind, I don't know).
> However, here is the real shocker - although the researchers were not
> looking for this, they discovered that 18-30% (different cities had
> different results) of the family samples had children with blood types
> which were so incompatible with the "father"'s sample as to firmly
> disprove paternity.

> I recently mentioned this to a researcher (though I do not think that he
> would object to my mentioning his name, I did not formally seek permission,
> hence I will not give his name out), whose response was that the statistic
> seemed rather high.  

(stuff deleted)

> At any rate, logical posturing aside, I am curious as to whether anyone has
> heard about the above study.  And if anyone has, does he/she have a journal
> reference of some sort?  I have made numerous requests before, and while I
> could simply ask the person who informed me about the study, I have been
> unable to contact him.

> Any help would be *greatly* appreciated.

I don't have a reference immediately at hand which addresses the issue of
the nonpaternity rate.  However, I have been working with the analysis end of
human pedigrees for many years, and I can say that the nonpaternity
rate is much higher than the average person would believe.  For example,
I have recently made a first pass at doing paternity testing on 22 pedigrees
which we have collected for a particular study, and of these, 4 had an
identifiable non-paternity.  (I insist on doing nonpaternity testing since for
many genetic studies, unidentified nonpaternity has highly deleterious effects
on the results of the analysis).  These aren't huge pedigrees - size 5-10
individuals in 2-3 generations.  One of these pedigrees had at least
3 different males involved in creating 5 children with one mother!
It is important to note, however, that the non-paternity rate per offspring
and the identifiable fraction of nuclear families with at least one 
nonpaternity are not the same thing.

The detectable nonpaternity rate is now much higher than it used to be.
The older blood group markers are inefficient for detecting nonpaternity
because they have few alleles.  The range of possible nonpaternity rates
you cite is outside of the detectable range with blood group markers, but
might be a plausible rate for highly polymorphic DNA markers.

Ellen Wijsman
Div of Medical Genetics, RG-25
and Dept of Biostatistics
University of Washington
Seattle, WA   98195
wijsman at u.washington.edu
(206) 543-8987

More information about the Gen-link mailing list