Establishing correlation between SNP and disorder

[geneTester22]

Hi,

As part of current research, I am doing RFLP on a SNP in the DYX1/2 loci on chromosome 6.

I am obtaining buccal cell samples for DNA, amplifying with PCR and then using restriction enzymes to test for prescence or absence of the SNP allele.

My question is how can I statistically analyze the data.

Say I have 3 homozygous wild-type, 5 heterozygous, and 7 homozygous affected.

What statistical testing can I do to establish a correlation between these data and a set of control data (the same RFLP run on a set of unaffected individuals)?

Thanks!

[GreenDog]

This is a really tough question, you've asked. this is a new area and not a very established one, you should search the pubmed you'll find some articles which have done this but I find them hard to understand. a girl in a lab next to me is working on the same problem, she has 5 SNP's...
Maybe you can use the hypergeometric distribution to calculate the chances? (Or was it Poisson distribution).
Anyway the best you can do is talk to your faculty statistician and ask his advice.

[wbla3335]

How about a Chi-square or Fisher's exact test on allele frequencies between your two populations.

[JorgeLobo]

Can you educate us re. SNP?

[GreenDog]

Chi-square wouldn't do, because as I understand the calculation needed is: What are the chances one will get the disorder if he's +/+, +/- or -/-. There are no expected ratios.

[wbla3335]

Perhaps the OP should clarify. I took the question as asking what test one could do to see if the SNP was associated with the disorder.