This sounds like an interesting project, but it is way too big and hard to
successfully tackle through email advice without the backing of an
understanding of the analytical methods and assumptions. This is an
example of a complex trait, and to have a chance of succeeding, what you
really need is a collaborator who is a statistical geneticist/genetic
epidemiologist. Such a person would be able to help you with design and
sample size requirements, as well as analysis and interpretation.
************************************************************************
Ellen M. Wijsman COURIER DELIVERY ADDRESS ONLY:
Research Professor Ellen M. Wijsman
Div. of Medical Genetics and 1914 N 34th St., suite 209
Dept. Biostatistics Seattle, WA 98103
BOX 357720, University of Washington (Note: Use this address
Seattle, WA 98195-7720 EXACTLY as given above, and
phone: (206) 543-8987 use ONLY for courier delivery!!!)
fax: (206) 616-1973 email: wijsman at u.washington.edu
web page: http://faculty.washington.edu/wijsman
*************************************************************************
On Mon, 19 Jul 2004, CHRIS wrote:
> Hi, there,
>> I am a new comer to the field of linkage analysis. I have couple
> questions regarding my ongoing project, and I would really appreciate
> your help.
>> I am a molecular biologist and I am designing an experiment to do a
> genome-wide mapping to identify certain loci (and hopefully genes
> eventually) that predispose an individual who is infected with
> Helicobacter Pylori to stomach cancer. In other words, it is well
> known that the H. Pylori causes stomach cancer. But certain people
> which specific mutations in certain unknown genes may be much more
> susceptible to the H. Pylori induced carcinogenesis. I collected about
> 20 extended families which each of them have more than three stomach
> cancer patients. And I tested the H. Pylori status for each of them
> (i.e., infected or not infected). Now my questions are:
>> One, how to calculate the sample size that will produce significant
> statistical power for the linkage analysis?
>> Two, I believe that I need to use the nonparametric model for
> analysis. Because there are two variates here, i.e., cancer status and
> H. Pylori infection status, how do I calculate the NPL Z score and the
> LOD score so that I can compare a certain locus between the patients
> who have cancer and are H. Pylori infected and the patients who have
> cancer but are not infected? Suppose I use MERLIN or GENEhunter
> programs.
>> I know these questions may sound stupid to you guys, but please help.
>> Thanks, Chris
>>
