Analyzing Imputed Data - How To Analyze A Specific Model ?

Question

[Dajewo]

Hello

i am suppose to do the gwas analysis and but because am not very familiar with it have a question about preparing data for mach2qtl.

the model i have to analyze is:

trait ~ SNP*age + SNP + age + sex

and i am not really sure how to get the SNP*age bit. Normally when the model is for example

trait ~ SNP + age + sex

i am simply doing linear regression against age and sex, getting the residuals which then am putting into mach2qtl.

don't know if anyone will be able to help me, but would really appreciate any help or tips how to do it.

Thanks

Answer 1

[Genotepes]

Hi Dajewo

I don't think MACH2QTL does SNP interaction. I knew of a version of SNPTEST but not sure it was freely available.

Obviously, you could just code this in R as the input format is easy to handle (dosage). But the GWA analysis could take quite a lot of time.

Plink now accepts dosage data and can do interaction with "hard" calls but I haven't checked whether it can do it's interaction model on dosage .. http://pngu.mgh.harvard.edu/~purcell/plink/download.shtml