I have genotyping data that i have already analyze of single SNP association with desired disease. For the further step, i need to analyze in haplotype scope (it mean i have to analyze association multiple SNP with desired disease). Actually, i have already try "haploview" platform, but it limited to look significant level of multiple SNP with desired disease. I don't know how to look at the Odd Ratio (OR) and 95% CI. Is there anyone know another program ? i really appreciate your response
Hi,
Forgive me if I am misunderstanding, but if your question is regarding addressing the association between the SNPs and a particular disease you could do this in many statistical packages. You could use something like SPSS (or mini-tab), import your variables (individuals identifier, SNP 1, SNP2 etc, and presence of the disease). You'd then need to ensure your variables are coded appropriately (ie are you simply coding people as a carrier or do you want to have different codes for hetero- and homozygous).
From there you could then perform a logistic regression analysis with disease as the dependent variable, and your SNPs as covariants. SPSS (and I imagine others) gives you the option to have odds ratios and their confidence intervals as an output, so you could select this (in SPSS, this meas selecting an option called "CI for exp(B)"). you will then get an odds ratio and CI for each SNP. The benefit of putting them all in the analysis together is you can control for confounding variables.
Hopefully that's not to basic an analysis,
Adam
of course Adam, I have already done that before. But the protocol as you mentioned above is the way if we look SNP as a single marker. Naturally,, there are multiple SNPs as a one unite that inherited together. In this step, we look these multiple SNPs as a "haplotype". Same as previous analysis, we need to look this haplotype (which is combination of SNPs) association with the disease. I'm just able to find out the significant level (haploview software provide this analysis), but i could not realize how to determine odd ratio and CI
Hi Aksar,
I am doing the same thing for my project. In my case, I used SPSS. You might put haplotype as a new variable in the variable view page. Then in the column of values, you define all different kinds of haplotypes possible to be coded as numbers. For Example, you choose value of "1" to label AA+GG haplotype, a value of "2" to label AA+GT, a value of no. "3" to label AA+TT, and so on until you put all possible haplotypes. Then you enter your Data for each case based on this coding. I am not sure why do you need to get odds ratio and CI. In my case, I just used Chi square as association between presence or absence of disease (association between two nominal variables).
Hope it helps,
Regards,
- Badges
- Science method