From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Wed Apr 16 2003 - 02:31:06 MDT
On Tue, 15 Apr 2003, Ramez Naam wrote:
> So I want to be able to determine the number of humans whose genotype
> and phenotype we'd have to capture in order to find this effect. More
> generally, I'd like to be able to express that necessary sample size
> as a function of the size of the effect we're looking for and the
> number of different genes involved.
Mez, you are dealing with the fundamental topic of "genotyping"
and gene "penetrance". Someplace I have a very thick file of
papers on these topics because we dealt with it at Aeiveos Sciences
Group for the Longevity Genotyping project.
The short answer is that you aren't going to get a good answer
(at least to the questions the way you have proposed them).
There are mathematical formulas that tell you how many individuals
you would need to genotype to identify the location of a gene
if you know how much of the "trait" the gene contributes and
how "penetrant" it is. For complex traits, e.g. longevity
(or intelligence) it is likely that the number of genes
contributing is high and the penetrance may be low -- so
you need a *lot* of genotypes to get statistical significance.
That isn't to say it is impossible -- but it is way beyond the
study scale that one requires to isolate common genetic diseases.
My gut feeling would be that you need thousands to tens of
thousands of individuals rather than hundreds (which is what
will usually suffice for diseases where there are founder
effects).
If you want to discuss this more, you may want to email me offlist
as its a fairly specialized topic.
Robert
This archive was generated by hypermail 2.1.5 : Wed Apr 16 2003 - 02:38:13 MDT